Collaborative Platforms: How they affect students' performance

Collaborative Platforms:

How they affect students’ performance

Inês Ferreira dos Santos Botelho Veiga

Dissertação de Mestrado em Gestão e Curadoria da

Informação

Fevereiro, 2022 Inês

Fe

rrei

ra d

os

San

tos

Bo

telh

o

Vei

ga, C

olla

bo

rati

ve P

latf

orm

s:

Ho

w t

hey

aff

ect

stu

den

ts’

per

form

ance

, 20

22

Collaborative Platforms:

How they affect students’ performance

Inês Ferreira dos Santos Botelho Veiga

A dissertation presented as a partial requirement for obtaining the master’s degree in

Information Management and Curation

NOVA Information Management School

NOVA School of Social Sciences and Humanities

Supervisor: Professor Doutor Carlos Tam Chuem Vai

2022

I dedicate this work to my family

Acknowledgements

I would like to thank Professor Carlos Tam and Professor Paula Ochôa for their precious

support and guidance.

I am very thankful to my family for motivating me, especially during the hard times.

I appreciate all students who made this investigation possible by providing me with their

valuable contributions and time.

Finally, a special thank you to my partner, whom I really appreciate for always helping me and

pushing me further.

Abstract

The topic of individual performance, in the context of technology in education, has

received limited attention. To study this matter, the Task-Technology Fit (TTF) theory

developed by Goodhue and Thompson (1995) was extended by other constructs found to

have an impact on university students’ performance, attending online classes.

Our mixed methods research sought to develop a model capable of predicting

students’ individual performance, by understanding the precursors of TTF, with the added

moderating role of facilitating conditions, and the precursors of utilization with the added

role of facilitating conditions. We sought to understand the role of voluntariness as a

moderator of the relationship between TTF and utilization, and, especially, the predictors of

individual performance, with the addition of two simple mediation models, with two

constructs each: environment and engagement, and perceived usefulness and compatibility.

The surveyed and interviewed students had a neutral to a somewhat positive

perspective of the impact of collaborative platforms on their individual performance. What

negatively impacted and thus weighed down this perception was, especially, the level of

engagement. Also, having a distracting environment with a bad internet connection

heightened the more negative level of engagement and thus contributed negatively to

students’ performance. Compatibility has an important mediating role, as perceived

usefulness, the strongest predictor of performance has an even stronger impact when

influenced by students' level of compatibility.

Interviews were conducted to explain and corroborate the significant impacts of

engagement, compatibility, and perceived usefulness on individual performance. They were

also important to explain the statistically insignificant impact of utilization on individual

performance and the impact of the surrounding environment on students’ engagement. They

also provided us with insight into the dichotomy of practical classes vs. theoretical classes,

the most useful functionalities of collaborative platforms, and future use. Even though

utilization was found to have no impact on performance we were able to provide further

insight into this complex variable. Finally, Voluntariness was found to be a very relevant

moderator and should be further tested in future research.

Keywords: collaborative platforms, online learning, individual performance, TTF

Resumo

No contexto da tecnologia na educação, o tópico do desempenho individual, tem recebido

pouca atenção. Para estudar esta questão, estendemos a teoria Task-Technology Fit (TTF),

desenvolvida por Goodhue e Thompson (1995), com outros construtos que pudessem ter

impacto no desempenho dos estudantes universitários que frequentam aulas online.

A nossa investigação de métodos mistos procurou desenvolver um modelo capaz de prever o

desempenho individual dos estudantes, ao entender os precursores de TTF com a adição do

moderador facilitating conditions, e os precursores de utilization com o papel adicional de

facilitating conditions. Procurámos compreender o papel de voluntariness como moderador

da relação entre TTF e utilization e, principalmente, os preditores de desempenho individual,

com a adição de dois modelos de mediação simples, com dois construtos cada: environment

e engagement, e perceived usefulness e compatibility.

Os alunos inquiridos e entrevistados denotaram uma perspetiva neutra a algo positiva sobre

o impacto das plataformas colaborativas no seu desempenho individual. O que impactou

negativamente e, portanto, denegriu essa perceção foi, principalmente, o nível de

engagement, que por sua vez, foi impactado negativamente por um ambiente com distrações

e uma má conexão à internet. Compatibility tem um papel mediador importante,

pois perceived usefulness, o preditor mais forte de desempenho, tem um impacto ainda mais

forte quando influenciada pelo nível de compatibility dos alunos.

As entrevistas foram realizadas para explicar e corroborar os impactos significativos de

engagement, compatibility e perceived usefulness no desempenho individual. Também foram

importantes para explicar o impacto estatisticamente insignificante de utilization no

desempenho individual e o impacto de environment no nível de engagement dos estudantes.

Também nos forneceram informações sobre a dicotomia aulas práticas vs. aulas teóricas, as

funcionalidades mais úteis das plataformas colaborativas e o uso futuro. Embora utilization

não tenha tido impacto no desempenho, fomos capazes de fornecer mais informações sobre

essa variável complexa. Por último, descobrimos que voluntariness é um moderador muito

relevante que deve ser explorado em investigações futuras.

Palavras-chave: plataformas colaborativas, aprendizagem online, desempenho individual, TTF

Abbreviations

AVE – Average variance extracted

COMP - Compatibility

EMS – Electronic meeting systems

ENG - Engagement

ENV – Environment

FC – Facilitating conditions

F2F – Face-to-face

ICTs – Information and communication technologies

INE – Instituto Nacional de Estatística

IT – Information Technology

LMS – Learning Management Systems

MIS – Management Information Systems

MOOCs – Massive open online courses

PBC – Perceived behavioral control

PEU – Perceived ease of use

PLS – Partial least square

PLS-SEM – Partial least square structural equation modeling

PU – Perceived usefulness

SEM – Structural equation modeling

TAM – Technology acceptance model

TASK – Task characteristics

TECH – Technology characteristics

TPB – Theory of planned behavior

TPC – Technology-to-performance chain model

TTF – Task-Technology Fit

UNESCO - United Nations Educational, Scientific and Cultural Organization

USE - Utilization

UTAUT – Unified theory of acceptance and use of technology

VLE – Virtual learning environment

VLNT - Voluntariness

WHO – World Health Organization

Contents

1. Introduction 1

2. Literature review 3

2.1. Context of the Study 3

2.2. Collaborative Platforms 7

2.3. Online Learning 10

2.4. Effectiveness of Online Learning 11

2.5. Emergency Online Learning 14

2.6 Task-technology Fit theory 19

2.7. TTF and Online Learning 21

3. Research Model 23

4. Research Methods 37

5. Quantitative Approach Design: Data Collection Methods 43

5.1. Surveys 43

5.2. Sample 43

5.3. Sample Characteristics 43

5.4. Data collection and measurements 45

6. Quantitative Approach Design: Data Analysis 48

7. Quantitative Results 50

7.1. Measurement Model 51

7.2. Structural Model 58

7.4. Explanatory Power 62

7.5. Moderation 63

7.6. Predictive Power 64

8. Qualitative Approach Design: Data Collection Methods 66

8.1. Interviews 66

8.2. Sample 67

8.3. Sample Characteristics 67

8.4. Data Collection and Measurements 68

9. Qualitative Approach Design: Data Analysis 69

10. Qualitative Results 69

10.1 Utilization 70

10.1.1. Platforms used, most used, utilization frequency, and Voluntariness 70

10.1.2. Utilization’s impact on individual performance 71

10.1.3. Future use 73

10.2. Facilitating Conditions 73

10.3. Technology Characteristics and Task Characteristics 75

10.4. Task-Technology Fit 78

10.5. Compatibility 81

10.6. Perceived Usefulness 84

10.7. Engagement 87

10.8. Environment 90

10.9. Individual Performance 93

11. Discussion 96

11.1. Technology Characteristics and Task Characteristics 97

11.2. Task-Technology Fit 99

11.3. Facilitating Conditions 100

11.4 Utilization 101

11.5. Perceived Usefulness 103

11.6. Compatibility 104

11.7. Environment 105

11.8. Engagement 106

11.9. Individual Performance 107

12. Implications 107

12.1. Theoretical Implications 107

12.2. Management Information Systems (MIS) Implications 110

12.3. Limitations and Future Research 111

13. Conclusion 112

References 114

List of Figures

Figure 1 – The subset of the Technology-to-Performance Chain model 20

Figure 2 – Research model with the proposed hypothesis 24

Figure 3 – Engagement mediation model 35

Figure 4 – Compatibility mediation model 36

Figure 5 – Methodological model for our mixed methods approach 42

Figure 6 – Structural model results 59

List of Tables

Table 1 – Sample characteristics 43

Table 2 – Items used to measure the constructs in our research model 45

Table 3 – Engagement’s indicators before and after removing the indicator ENG3 52

Table 4 – Indicators before and after removing the items with the lowest outer loading 53

Table 5 – Indicator’s loadings and cross-loadings 55

Table 6 – Heterotrait-monotrait ratio 57

Table 7 – Total effects and direct effects of the significant indirect paths in our model 61

Table 8 – 𝑓2 of the significant paths in our model 62

Table 9 – The moderators’ 𝑓2 64

Table 10 – 𝑃𝐿𝑆𝑝𝑟𝑒𝑑𝑖𝑐𝑡 results 65

Table 11 – Interviewed students characteristics 67

Table 12 – Questionnaire responses to utilization and voluntariness’s items 70

Table 13 – Questionnaire responses to facilitating conditions’ items 74

Table 14 – Questionnaire responses to technology characteristics’ items 77

Table 15 – Questionnaire responses to task characteristics’ items 78

Table 16 – Questionnaire responses to TTF ’s items 80

Table 17 – Questionnaire responses to compatibility’s items 83

Table 18 – Questionnaire responses to perceived usefulness’s items 86

Table 19 – Questionnaire responses to engagement’s items 89

Table 20 – Questionnaire responses to environment’s items 92

Table 21 – Questionnaire responses to individual performance’s items 95

List of Graphics

Graphic 1 – Moderating Effect of FC on the relationship between TECH and TTF 108

Graphic 2 – Moderating Effect of VLNT on the relationship between TTF and USE 109

1

1. Introduction

Collaborative platforms are an important asset for facilitating communication among

individuals, information sharing, and knowledge co-creation (Cheung & Vogel, 2013; Massey,

2008). They enable distant individuals and help create opportunities for businesses and

organizations (Bélanger & Allport, 2008; Pavlou et al., 2008). They can reduce costs and

improve effectiveness, as time and place constitute no constraints (Dennis et al., 1988;

Meroño‐Cerdan et al., 2008). They enable researchers and specialists to connect and build

knowledge (Lundvoll Nilsen, 2011), employees to telework (Massey, 2008), and organizations

to innovate (Michaelides et al., 2013). They improve education quality for distance and

blended courses (Francescucci & Foster, 2013; Francescucci & Rohani, 2019; Zawacki-Richter

& Naidu, 2016). They help students who need to attend classes in crisis situations (Rajab,

2018), students who live in countries where access to schools and universities is difficult

(Shakya et al., 2017), students with special needs (Baykal et al., 2020), and students who are

parents and workers, and so need more freedom in their time management (Huda, 2011;

Murphy & Yum, 1998).

Collaborative platforms originated in 1985 (Dennis et al., 1988; Pavlou et al., 2008) and

have improved their utility and effectiveness since the surge of wireless networks and

portable devices (Bélanger & Allport, 2008; Lewin et al., 2018). They have been primarily used

by organizations, however, when access to this IT improved, schools and universities started

to take advantage, especially for online learning (Brodahl & Hansen, 2014; Massey, 2008;

Yadegaridehkordi et al., 2019).

Currently, education institutions are incorporating Zoom, Google Tools, and other

online collaborative platforms in their course design, as they enable collaborative work during

classes and help students to collaborate after classes (Hidayanto & Setyady, 2014; C. W. Taylor

& Hunsinger, 2011).

The study of collaborative platforms, in the context of education, still has a long way

to go, as they have only been recently adopted, due to the improvement in accessibility and

lowering in costs, and due to the internet’s universalization and the use of portable devices

(Jeong et al., 2019; Zawacki-Richter & Naidu, 2016).

2

There have been studies on the use of technology for collaborative learning (Jeong et

al., 2019; Laal & Laal, 2012; Resta & Laferrière, 2007). In fact, the term collaborative learning

is widely used and well established in education research (Jeong et al., 2019). However, the

focus of these studies is the teaching and learning method, which involves groups of learners

working together to create an outcome (Laal & Laal, 2012), and Internet-based technologies

that promote social interaction, cooperation, and collaboration, for learning and knowledge

building (Resta & Laferrière, 2007).

When it comes to the term collaborative platforms, it is rarely used in education

research, and the studies mainly focus on use and adoption (Cheung & Vogel, 2013;

Yadegaridehkordi et al., 2019), and only a few focus on how collaborative platforms can affect

students’ performance (Laakso et al., 2018; Orme et al., 2020). Also, we have noticed that the

term collaborative platforms, gets lost in education research. More recently, due to the

COVID-19 pandemic, there has been an increase in studies related to collaborative platforms

for online learning. However, the term collaborative platforms, is not used in these studies

(Flores et al., 2021; Soria et al., 2020).

We propose and test our own theoretical model based on Goodhue and Thompson’s

(1995) TTF theory. However, our model explores beyond the impact of utilization and TTF on

individual performance. We introduce two moderators to the traditional model and add two

simple mediation models to further explain students’ individual performance.

In the studies about online learning technologies, where a theoretical model based on

the task-technology fit (TTF) theory (Goodhue & Thompson, 1995) is established and tested,

there have not been many attempts in understanding what factors, beyond Utilization and

TTF (Bere, 2018; McGill & Hobbs, 2008), affect individual performance. In fact, there are

studies, in the context of online learning, that use the TTF theory but do not approach

individual performance (Pal & Patra, 2020; Wu & Chen, 2017).

These factors highlight how important it is to add studies on students’ individual

performance to Information Management Systems’ body of knowledge. In our study, we

propose to study collaborative platforms’ impact on university students’ individual

performance, to discover what factors affect Portuguese students using collaborative

platforms for online learning, in terms of their individual performance.

3

To study this matter, we propose a sequential mixed methods research, where a

survey is going to be employed to provide answers to our hypothesis, and interviews are going

to be conducted to explain the research questions that arose from the quantitative study,

provide breadth and depth to our findings, and help to corroborate and converge them.

2. Literature review

2.1. Context of the Study

On the 31st of December 2019, China reported, to the World Health Organization

(WHO), the first cases of COVID-19 (WHO, 2021). The virus started to spread quickly, and a

month later, there were 98 cases in 18 countries outside of China (WHO, 2021). On the 11th

of March 2020, WHO characterized the COVID-19 outbreaks as a global pandemic, since there

were more than 118 000 cases and 4291 deaths reported in 114 countries (Cucinotta &

Vanelli, 2020; WHO, 2021). Around this time, Europe was the most affected region,

accounting for more than 40% of the globally confirmed cases (WHO, 2021). Portugal, by the

end of March, reached 1305 cases of COVID-19 and 24 deaths (Dong et al., 2020).

The rapid spread of the virus led governments to implement partial and full lockdown

measures and by Abril more than 90 countries and half of the world population was affected

by these measures (Fahey, 2020; Sandford, 2020). This meant education was going to be

severely affected all around the globe. According to UNESCO, on the 1st of April, the number

of schools closed and partially open due to the COVID-19 virus was at the highest, with more

than 1.713 billion affected learners (UNESCO, 2020b). This number represents about 1/7 of

the world population and more than 91% of the total enrolled learners in the world (UNESCO,

2020b).

In Portugal, more specifically in the capital Lisbon, on the 9th of March 2020, the

University of Lisbon was the first teaching institution to close its facilities and implement

distance learning, due to the COVID-19 pandemic (Reitoria da ULisboa, 2020). On the 16th of

march 2020, the ministry of education suspended all face-to-face school activities which led

to schools’ closures, in the entire country (XXII Governo, 2020). Following the lead, on the

same date, NOVA University suspended all of its face-to-face classes (Universidade NOVA de

Lisboa, 2020a). Ultimately, on the 19th of March, the government declared the first emergency

4

state, on the basis of a public calamity situation (Presidente da República Portuguesa, 2020),

which meant that Portuguese inhabitants were obligated to be confined to their homes.

Students from primary, secondary, and tertiary education began to have classes online, and

non-essential workers started to work from home.

On the 30th of April 2020, Portugal’s deconfinement phase started, which led to

schools and universities’ partial reopening on the 18th of May 2020 (Presidência do Conselho

de Ministros, 2020), until they were fully open after the summer break, which ended on the

13th of September 2020. But the nightmare was not over, and on the 15th of January 2021,

the mandatory confinement started again (2021d). Despite efforts to maintain schools and

universities open, by introducing a 15-day academic break starting on the 22nd of January

2021, on the 8th of February 2021 schools and universities were shut down again, due to the

COVID-19 virus (2021a; Universidade NOVA de Lisboa, 2021). On the 15th of March 2021

schools started to partially reopen (Conselho de Ministros, 2021), and on the 19th of April

2021, everything came closer to normality with universities and schools fully reopening in the

entire country (2021b, 2021c).

During the COVID-19 pandemic, schools, and universities, in an attempt to best

replace in-person classes, transitioned to an online learning environment and recurred to

online tools, which facilitated communication and group work. For instance, NOVA University

relied on Zoom and Google Hangouts Meet for synchronous online classes; and Nonio and

Moodle for sharing learning resources, submitting papers, conducting evaluation quizzes, and

tests (Universidade NOVA de Lisboa, 2020b). Likewise, the University of Lisbon relied on Zoom

for synchronous online classes, FCCN EDUCast for asynchronous online classes, Moodle for

sharing educational content, and Teamviewer for remote collaborative group work (ULisboa,

2020).

Collaborative applications and platforms, communication, and conference tools, as

well as collaborative document editing, have all made the transition for schools, and

universities easier. They provide a variety of learning possibilities and solutions for group

projects and promote consistent participation, prompt communication, regular group

discussion, timely and relevant contributions, and commitment to the task (Carrillo & Flores,

2020). Since the COVID-19 pandemic started, the number of users of collaborative platforms

has grown. For instance, Zoom, in April of 2020, surpassed 300 million daily meeting

5

participants (Zoom, 2020), which coincides with the month with the most affected learners

and country-wide schools’ and universities’ closures (UNESCO, 2020b). Microsoft Teams, in

the same month, had more than 200 million daily meeting participants (Windows Central,

2020) and Google Meet over 100 million (Peters, 2020).

Zoom was one of the most used collaborative platforms in higher education

institutions for remote classes (Vargo et al., 2021). Therefore, it is important to acknowledge

its functionalities. Zoom is a video conferencing platform that supports live HD video and

audio transmission and it is compatible with several devices, such as portable computers and

tablets (Zoom, 2021). Zoom allows sessions’ recording, automatic transcription of classes, and

it integrates with learning management systems (LMS), like Moodle (Zoom, 2021).

Furthermore, it has suitable functionalities for online classes such as a polling tool, a group

chat, one-click content sharing, real-time co-annotations, digital whiteboarding, screen

sharing, and video breakout rooms (Zoom, 2021).

Despite Zoom having many adequate functionalities for online learning, the transition

from face-to-face learning to remote learning was sudden and involuntary so, some

difficulties naturally emerged, such as poor online teaching infrastructures, the inexperience

of teachers, learning gaps, the environment at home, and the lack of competencies in the use

of digital instructional formats (Carrillo & Flores, 2020). Students with access to technology

and the internet at home had a smooth transition, however, families with fewer resources

saw a hard time in filling learning gaps and in balancing work obligations with childcare

(UNESCO, 2020a).

The digital divide has a big part in the sufficient use of online collaborative platforms

(Koranteng et al., 2020; UNESCO, 2020b). To use collaborative platforms online, one needs

access to the internet and a device, such as a computer or a smartphone. Therefore, for

contextualizing this study, it is important to assess Portuguese students’ digital divide.

According to Instituto Nacional de Estatística (INE), in 2020, when compared to 2019,

there was a significant increase in internet access attributed to the mandatory confinement

during the COVID-19 pandemic, where people had to work and study from home (INE, 2020).

In 2020 and 2021, 84,5% and 87,3%, respectively, of the Portuguese families had access to

the internet at home, which constitutes a significant increase when compared to 80.9% in

6

2019 (INE, 2020, 2021). Likewise, in 2020 and 2021, 81,7% and 84.1%, respectively, had

broadband internet, a rise from the 78% in 2019 (INE, 2020, 2021). When it comes to the ages

targeted in our study, 99.5% and 99.7% of Portuguese users between the ages of 16 and 24

and 98.2% and 98.4% of Portuguese users between the ages of 25 and 34 used the Internet

in 2020 and 2021, respectively (INE, 2020, 2021). This percentage is higher when we focus on

students, with 99.8% of Portuguese students using the Internet in 2020 and 2021 (99.6% in

2019) (INE, 2019, 2020, 2021). This should mean that Portuguese university students, as in

other developed countries, would have the necessary resources for the sufficient use of

collaborative platforms.

Students in developed countries have greater exposure to technology and are more

prone to learn digital skills (Steele, 2019). When it comes to communication-related activities,

89.9% and 91.4% of the Portuguese population used the internet to send instant messages,

86.8% and 87.6% used the internet to send or receive emails, and 70.5% and 79.7% used the

internet for calls or video calls, in 2020 and 2021, respectively (INE, 2021). Not only did

internet use increase for communication and work purposes, but also for the use of e-

commerce (INE, 2020, 2021). Furthermore, in Portugal, the internet’s utilization and e-

commerce use decreases when the age group increases, and students are the predominant

group using both, even surpassing the employed (INE, 2020, 2021). Also, INE’s 2020 and 2021

inquiry indicates that the higher the level of education the higher the use of internet and e-

commerce. This should mean that in Portugal, as in other developed countries, students are

prone to innovation and have a high level of digital skills, therefore, they will adapt well to

the use of a new software, such as collaborative platforms, for online learning.

In conclusion, even though the change in education delivery was abrupt and a novelty

for most Portuguese students, the transition probably felt easy for university students. As

these students have a high level of education, are familiar with the online world, have a high

level of digital skills, and access to a good internet connection, they should have had an easier

transition from a traditional learning method to an online one, and should offer less resistance

to the use of a new technology.

7

2.2. Collaborative Platforms

Collaborative platforms allow a group of people to work on the same task and

communicate simultaneously from different locations. They were designed to improve group

productivity, no matter its size, and can be used for synchronous and asynchronous meetings,

collocated and remote meetings (Bîzoi et al., 2009). These collaborative IT tools have been

studied primarily in the context of organizations and have been established as important tools

for achieving effective collaborative work (e.g. Ayala et al., 2020; Marion & Fixson, 2021;

Qaddumi et al., 2021)

Electronic Meeting Systems (EMS) were the first version of the IT tools we currently

use for collaborative group work (Pavlou et al., 2008). PLEXSYS, an Information technology to

support meetings, was introduced in the University of Arizona, in 1985, with a focus on

communication (Dennis et al., 1988). Compared to group decision support systems (GDSS)

and computer-based systems for cooperative work (CSCW), EMS went a step further by

allowing communication across time and space for large and small groups, facilitating

decisions, projects' completion, and other tasks (Dennis et al., 1988). EMS were created to

facilitate a group's task performance and enhance communication among individuals.

PLEXSYS was implemented in a multinational corporation, and in a year, it was able to improve

the organization's productivity with a decrease in the number and duration of meetings

(Dennis et al., 1988). On the other end, meetings not supported by PLEXSYS were less

productive and projects would take significantly more time to be completed (Dennis et al.,

1988). Also, the size of the group attending the meeting had no impact on the productivity of

the meeting, plus it facilitated the participation of individuals from different hierarchical levels

improving organizational communication and decision making (Dennis et al., 1988).

The Internet, wireless networks, high bandwidth networks, mobile phones, and

portable computers, triggered a change in the paradigm, as they facilitated and extended the

use of collaborative IT tools for real time information sharing and communication, among

distributed individuals, anytime, and anywhere (Bélanger & Allport, 2008; Pavlou et al., 2008),

weather from an office space, the comfort of someone’s home, or when walking the dog.

Collaborative platforms are especially important in an uncertain environment, as they will be

more important for a positive performance (Pavlou et al., 2008). Environmental uncertainty

can be defined as all the external factors that are uncontrolled by the group and can suffer

8

unexpected changes, and in an uncertain environment, communication and information

sharing enable groups to better respond to the environment (Pavlou et al., 2008). This has

especially proved to be true as collaborative platforms were very important to help navigate

the uncertain environment created by the COVID-19 virus.

In the context of education, collaborative platforms have already been approached in

a small number of studies (Brodahl & Hansen, 2014; Dishaw et al., 2013; Hidayanto & Setyady,

2014; Laakso et al., 2018; Yadegaridehkordi et al., 2019), and they all focus primarily on

adoption and utilization.

Taylor and Hunsinger (2011), in a mixed methods study, applied the theory of planned

behavior (TPB) to study what factors influenced university students’ intention to use Google

Docs, a collaborative tool. The authors concluded that students’ intention to use Google Docs

was directly predicted by Attitude, Subjective Norms, Perceived Behavioral Control, and

Affect. For obtaining data, they conducted 15 interviews and 316 surveys at a Canadian

business college.

Cheung and Vogel (2013) studied Google Applications’ user acceptance, as an example

of web based collaborative technologies, in the context of e-learning, in an advanced

Marketing research course from a Hong Kong university. The tools included in their study

were Google Docs, Google Forms, Google Sites, Google Group Forums, and Google Drive, and

they were used in the context of group projects that required collaboration (Cheung & Vogel,

2013). To understand students’ adoption behaviors, Cheung and Vogel (2013), extended the

Technology Acceptance Model (TAM) and TPB. They concluded that the key factors that

influenced students’ intention to use Google Applications, were Attitude, Peer influence, and

Perceived Ease of Use, which is strongly predicted by compatibility. To collect data 136

questionnaires were conducted.

Dishaw et al. (2013) made a comparative study with four collaborative writing and

editing technologies, where American university students attending a business major had to

collaboratively write a research paper. The technologies in analysis were MS Word combined

with email, Twiki, Google Docs, and Office Live and they were studied through a model that

combined TAM and TTF, to measure Task-Technology Fit (TTF), Perceived Ease of Use (PEU),

9

and Perceived Usefulness (PU). The authors conclude that Google Docs had a good TTF, PEU,

and PU. The study included 1002 participants.

Hidayanto and Setyady (2014) studied the use of collaborative tools for completing

group projects in the Faculty of Computer Science of Universitas Indonesia. The authors

investigated collaborative platforms’ adoption factors and their impact on group performance

and for that, they extended the TAM model. They concluded that the use of collaborative

platforms had a positive significant impact on group performance. The authors obtained 196

questionnaire responses.

Brodahl & Hansen (2014) presented a case study on education students’ perceptions

of collaborative writing using Google Docs and EtherPad. The authors concluded that only a

minority of students felt the quality of group collaboration increased with use, and for most

students, the tool did not work as expected. However, they felt positive about the

functionality of commenting and editing text collaboratively, but when group sizes were large

that could be a frustrating experience. They collected 154 questionnaires and 145 reflection

notes.

Orehovački and Babić (2014) did research on higher education students’ continuance

intentions regarding the use of Collaborative Web 2.0 applications, more specifically Google

Docs. The authors combined and adapted the TAM model with the Expectation-Confirmation

Theory (ECT). The students from this study were enrolled in the course “Foundations of

Informatics” from Polytechnic of Rijeka and had to develop a syllabus based on a topic, as part

of a group project. They concluded that Satisfaction and PU were predictors of Continuance

Intention. The data was collected from 127 questionnaires.

Laakso et al. (2018) presented in their paper four case studies where ViLLE, a

collaborative learning tool created by them, was implemented successfully. Two of those case

studies were conducted at the University of Turku. The authors concluded that ViLLE had a

positive impact on learning, as it offered a practice environment with interactive exercises

with the ability to engage students (Laakso et al., 2018).

Yadegaridehkordi et al. (2019), in the context of higher education, approached the

adoption of cloud computing, a collaborative technology, in four top Malaysian universities.

To understand the adoption of online collaborative learning tools, Yadegaridehkordi et al.

10

(2019) extended the TAM model. They concluded that PU was a predictor of Intention to

Adopt, which was in turn predicted by collaboration (i.e., features that facilitate students’

collaboration). The study was based on data from 209 questionnaires.

Orme et al. (2020) studied the impact of Microsoft Teams on group work, in a blended

course about research methods. The authors concluded that using Microsoft Teams can

benefit students’ performance and the experience of working in groups. In fact, they

considered this experience resulted in the most positive outcomes they had ever seen, as

students’ work improved in quality with more creative and challenging ideas and projects.

From the literature reviewed we can conclude that the studies on online collaborative

platforms, in the context of higher education, are recent and focus mainly on utilization.

Furthermore, there has not been one study that tests and builds a model to predict students’

individual performance, when utilizing a collaborative platform. Therefore, there is a lot of

room for testing different theories and exploring a performance-centered model.

Nonetheless, we noticed the term collaborative platforms is not commonly used in the

context of education. Being so, we were able to identify more studies that approach

collaborative platforms but do not label them as so. This has especially happened in studies

about collaborative platforms used during the COVID-19 pandemic, which are going to be

approached in section 2.5. This uncertain situation has opened the eyes of researchers,

institutions, and teachers for the impact these platforms can have in students' education.

2.3. Online Learning

Between 1985 and 1989 educational technologies and media were introduced as a

way of improving quality, equity, and participation in distance education (Zawacki-Richter &

Naidu, 2016). These technologies started to be seen as an answer to the students' and

teachers’ needs, and as a way to combat the limitations and challenges imposed by the

separation of learners, teachers, and the teaching institutions (Zawacki-Richter & Naidu,

2016). Among these technologies were: audio teleconferencing, interactive telecourses, and

electronic mail, but they did not offer a synchronous experience that closely resembled F2F

classroom interactions (Francescucci & Foster, 2013; Zawacki-Richter & Naidu, 2016). Since

then, there has been an investment in newer and better technologies to support a more

11

efficient two-way interaction between teachers and students (Zawacki-Richter & Naidu,

2016).

The end of the XX century marked the early stage of online learning and the beginning

of virtual universities, with the development of the first Learning Management Systems (LMS),

which were primarily used for uploading learning content (Singh & Thurman, 2019). At the

time, online learning reduced time and place constraints and allowed students to participate

from their convenient location using a computer and a user-friendly and customizable online

learning program (Francescucci & Rohani, 2019).

In the XXI century, online learning and virtual universities minimize the differences

between distance education and traditional face-to-face (F2F) education because of the new

technologies that enable synchronous communication (Francescucci & Foster, 2013;

Francescucci & Rohani, 2019; Gurcan & Cagiltay, 2020; Zawacki-Richter & Naidu, 2016). Many

universities saw the potential of the new Information and Communications Technologies

(ICTs), as they facilitated collaborative online learning and teaching, at a reduced cost, and

with greater flexibility and convenience. Therefore, they started to offer blended learning,

which is a mix of face-to-face and distance learning (Francescucci & Foster, 2013; Zawacki-

Richter & Naidu, 2016). Currently, students learn online in a flexible, synchronous, and

comfortable environment with the help of digital tools and ICTs, which enable students to

interact online and collaborate with other students and teachers, in an environment that

closely resembles the experience of a face-to-face classroom interaction (Francescucci &

Foster, 2013; Gurcan & Cagiltay, 2020).

In the next decade, distance learning’s focus will continue to be on the design of

distance learning platforms adaptable to the specific needs of individuals, which will

potentially lead to the creation of intelligent distance education platforms (Gurcan & Cagiltay,

2020).

2.4. Effectiveness of Online Learning

Most of the literature about online learning in higher education focuses on use and

adoption (Khechine et al., 2020), on students' perceptions and attitudes towards online

learning (Al-Salman & Haider, 2021; Rojabi, 2020; Serhan, 2020), on students’ online learning

readiness (Chung et al., 2020; Joosten & Cusatis, 2020; Latheef et al., 2021) and on students’

12

satisfaction (Zeng & Wang, 2021), but also on performance (T. Davis & Frederick, 2020; Wei

& Chou, 2020).

Since the 1980s, for online learning, the key has always been collaboration and

communication facilitated by technology, as it brings teachers and students together,

minimizes the differences between F2F and online learning, and offers a way of

complementing traditional F2F learning. Online learning is the latest form of distance learning

and it has many benefits such as: no time constraints, freedom in time management, no place

constraints, easy access to the course's materials, asynchronous and synchronous

interactions, group collaboration, and cost reduction (Liaw, 2008). However, it can have some

downsides, such as the lack of a firm framework to encourage students to learn, the students’

need to have a high level of self-discipline, and it can feel lonely and demotivating, because

there is a lack of a learning atmosphere and a lack of interpersonal and direct contact and

discussion among students and teachers (Liaw, 2008).

Webster and Hackley (1997) conducted a study on technology-mediated distance

learning outcomes by focusing on technology characteristics and concluded that perceived

medium richness had an impact on students’ learning outcomes. Therefore, they advised

instructors to focus on the varied functionalities the technology can offer. Furthermore,

students’ comfort with their images on screen, instructor control over the technology, the

quality of the technology, instructors’ attitudes, and teaching style were found to have an

impact on students’ performance. Also, the teaching style influences involvement and

participation, and an interactive teaching style that promotes discussion and interaction

among students and with the professors is advised. Furthermore, Webster and Hackley (1997)

advised the instructors to help students’ feel comfortable, learn how to utilize the technology,

and be positive, for the best learning outcomes. Likewise, Liaw (2008) concluded that learning

content that uses different multimedia is essential for improving e-learning efficacy,

performance, and individual's motivation. Similarly, Davis and Frederick (2020) found that

online courses designed with richer multimedia could reduce students’ cognitive load,

increase engagement, and thus increase performance.

In a revision of the literature related to reports of performance outcomes and

effectiveness of distance learning, Zhao et al. (2005) found that studies published in and after

13

1998 reported that distance education was significantly more effective than face-to-face

education. This is due to the rapid evolution of learning technologies that become more

adequate for quality learning (Y. Zhao et al., 2005). It is also important to note that when

students were the evaluators of the learning outcomes, face-to-face was better than distance

education but the difference was not significant (Y. Zhao et al., 2005). Also, when the

instructor was more involved, distance education had significantly better outcomes when

compared to face-to-face education (Y. Zhao et al., 2005). Finally, distance education courses

that included forms of face-to-face interaction and combined synchronous communication

with asynchronous communication had better learning outcomes (Y. Zhao et al., 2005). The

authors concluded that there was a need for a live present human instructor for effective and

quality distance education, and video conferencing was suggested as an effective tool for

face-to-face communication (Y. Zhao et al., 2005).

Asynchronous distance communication is defined as two or more individuals

communicating via a common medium at different times and places (Alanazi, 2019). Online

courses that focus on asynchronous communication have their advantages and

disadvantages. They are more convenient for the learner since students can learn at their own

pace and preferred schedule, independent of other learners (Alanazi, 2019). It also gives

learners more time to reflect since they can pause and rewatch the lectures (Alanazi, 2019),

therefore there is a deeper engagement with learning contents (Watts, 2016). On the other

side, communication through asynchronous channels can generate confusion and there is no

immediate interaction or response from instructors and classmates (Alanazi, 2019). Students

can feel demotivated and isolated because there is a lack of social presence and interaction

with professors and colleagues, who can appear disconnected, which results in poorer

learning outcomes, lower retention rates, and higher dropout rates (Alanazi, 2019). That is

why students feel like tools for interacting and communicating in learning activities are the

most important for system quality, which is in turn important for e-learning effectiveness

(Liaw, 2008)

Synchronous distance communication occurs when two or more individuals

communicate through the same medium at the same time but in different places (Alanazi,

2019). It includes textual tools like Microsoft Teams’ chat box, as well as media tools such as

Zoom’s screen share (Alanazi, 2019). Online courses with synchronous communication allow

14

for immediate feedback and interaction, increasing students' social presence (Alanazi, 2019).

Students tend to feel more engaged as there is instantaneous feedback, also they can see the

faces of their colleagues and professors and interact with them, creating the sense of a deeper

connection (Watts, 2016). However, it can have some obstacles like matching schedules with

every participant, scheduling with different time zones, and the large dependence on the

internet connection's speed (Alanazi, 2019; Watts, 2016).

Online courses usually contain synchronous features, to emulate F2F classes, such as

instant messaging and videoconferencing, plus asynchronous features, such as content

storage and retrieval, email communication, document drop boxes, grade inquiries, and

learning objects (Francescucci & Foster, 2013). When it comes to distance and blended

learning, which is a mix of F2F and distance learning, collaborative, and interactive learning,

with a live instructor and peers, has been proven to lead to the best performance outcomes

(Francescucci & Foster, 2013; Francescucci & Rohani, 2019). Therefore, for e-learning

effectiveness and satisfaction, it is important to have synchronous and asynchronous

interactions (Liaw, 2008; Liaw & Huang, 2013).

Human interactions have been proven to be a fundamental construct for students'

satisfaction, course quality, and better learning outcomes (Alanazi et al., 2019; Francescucci

& Foster, 2013; Liaw, 2008; Watts, 2016; Y. Zhao et al., 2005). Since Google Tools, Microsoft

Teams, and Zoom are all collaborative platforms focused on synchronous communication

with live video, sound, and text, they can be a great asset to promote student-student

interactions and student-teacher interactions, and therefore positively influence students’

individual performance.

2.5. Emergency Online learning

Our study is set in Portugal, in a specific period in time, where students were forced

to have all of their classes, group reunions, and most of their evaluation elements, remotely,

due to a crisis situation (Flores et al., 2021). These students were enrolled in a traditional face-

to-face course but, because of the COVID-19 virus, all the scheduled classes had to migrate to

an online setting.

Therefore, we must note that planned online distance education is not the same as

emergency online distance education. It is a temporary and sudden shift in the delivery of F2F

15

education due to crises such as war, weather, and health (Hodges et al., 2020; Moser et al.,

2021). And COVID-19 was a worldwide health crisis, with an impact on students from all

sectors of education, teachers, and parents. Almost everyone was forced to work and study

from home, for an extended period, and although distance education has been linked to

positive learning outcomes, these students had no choice and institutions were not prepared

to embark on online learning. In this section, we are reflecting on some of the studies

conducted in the context of crisis situations.

Rajab (2018) conducted a study on Najran University, in Saudi Arabia, which had to

suspend its F2F classes and adopt online learning, due to a war between Saudi Arabia, the

Arab Coalition, and Yemeni rebel groups. Rajab (2018) found no difference in students'

performance when comparing online education and F2F education. This is because online

learning offers a safe learning environment, has the potential for delivering quality education,

and has engaging mediums, so it is a great option for delivering education in zones affected

by crisis (Rajab, 2018). Najran University used Blackboard, a Learning Management System

(LMS), to transfer all the courses to an online environment, and it was the first time something

like this happened in Saudi Arabia. Overall, online learning and F2F education had a similar

impact on students' performance, as both had similar passing and dropout rates, there was a

high level of interactivity between students and between students and teachers, the

instructors were actively engaged with the resources available for teaching online, there was

a high percentage of participation, low percentage of missed classes, moderate number of

technical issues, and high rates of course accomplishment (Rajab, 2018). The study results

were formed based on objective data, so objectively there were no differences in

performance for both learning environments. However, if students' perception was

evaluated, the results might have been different. For example, the enrolment rates in the

online semester were significantly lower than the F2F semester, this could be for the lack of

confidence in online learning effectiveness since that is the educators' and policymakers’

perception in Saudi Arabia (Rajab, 2018).

Serhan (2020) conducted a study where, due to COVID-19, 31 college students from

the USA, had to stop attending on-campus F2F lectures and switch to remote classes via

Zoom. Participants of the questionnaire had an overall negative attitude towards the use of

Zoom, with only 23% of the students enjoying using Zoom (Serhan, 2020). Following the same

16

trend, students believed that Zoom had a negative impact on their learning, with only 10% of

the surveyed students considering that the use of Zoom improved their learning (Serhan,

2020). When it comes to students' engagement, students believed Zoom had a negative

impact on their engagement, as it did not promote interaction, with only 13% of students

considering that Zoom helped them to participate in classes (Serhan, 2020). Students also

compared the two types of learning, and the results showed they preferred F2F learning

(Serhan, 2020).

In the same study, students were asked to identify Zoom's advantages and

disadvantages. In terms of the advantages, 79% of students considered the platform very

flexible, as it met their scheduling needs, and they could do it from anywhere. Also, being in

the comfort of their home, as there was no need to go to a physical location, and, finally, not

being required to show their faces (Serhan, 2020). 11% of students considered Zoom to allow

easier interaction, especially with the teacher, because it allowed them to ask questions in

real-time (Serhan, 2020). Furthermore, 5% of the students saw the possibility of written

communication as a benefit because it removed the need for public speaking, the same for

the use of multimedia (Serhan, 2020).

When it comes to the disadvantages, the consensus for 42% of the students was the

distractions in the home environment, as in their home, outside of a physical classroom, there

were more distractions in the environment, such as the phone or family members, which

made focusing harder (Serhan, 2020). 37% of students think there is a lack of quality in the

interaction and feedback they get from the instructor and other students, 16% considered

they were getting a poor education via Zoom, not understanding the content taught and

losing motivation, and 5% experiencing technical difficulties (Serhan, 2020).

Lau and Sim (2020) conducted a study at the School of Science and Technology (SST),

in Malaysia, where F2F education had to be replaced by online education, due to the COVID-

19 pandemic. To reach the results, Lau and Sim (2020) questioned more than 400 students

for each semester of the year of 2020, using qualitative and quantitative questions. In March,

for the first semester, 85% of students had an ok to positive experience, and in August, for

the second semester, that percentage rose to 91% (Lau & Sim, 2020). Furthermore, it is

interesting to note that in the first semester most students (42%) had an ok learning

experience, but in the second semester most students (45%) had a good learning experience

17

(Lau & Sim, 2020). When it comes to quality of teaching delivery, quality of learning materials,

which is the dimension with the most positive perceptions, and communication with

lecturers, most students had a positive outlook, which grew even more positive in the second

semester (Lau & Sim, 2020). However, students' perception of communication amongst

classmates got the least positive results, with about ⅓ of students having a bad experience

for the first semester, however, the negative perception reduced to half in the second

semester (Lau & Sim, 2020).

Students most appreciated the flexibility, convenience, and freedom of online classes,

and they considered it an easy and enjoyable experience, as it was from the comfort of their

homes (Lau & Sim, 2020). Students felt difficulties in following or focusing on what was being

thought and experienced an unstable internet connection (Lau & Sim, 2020). Furthermore,

they felt distracted, confused, and procrastinated more (Lau & Sim, 2020). Finally, students

prefer to have F2F classes but out of 483 students, 172 would prefer to continue with online

classes instead of F2F classes (Lau & Sim, 2020).

Soria et al. (2020), made an investigation on the obstacles to remote learning for

university students, and concluded that the biggest obstacle among students at all levels was

the lack of motivation for remote learning, followed by lack of interaction or communication

with other students. But also, the inability to learn effectively in an online format, and a

distracting home environment or lack of access to an appropriate study space (Soria et al.,

2020).

Flores et al. (2021), during the first year of the COVID-19 pandemic, conducted a study

with 2718 students enrolled in Portuguese universities across the country. The students’

opinions on their adaptation to online learning was very divided with 39% having a negative

perception, due to difficulties in concentrating for being at home, experiencing stress and

anxiety, experiencing a loss in education quality attributed to a lack of practical classes,

suspension of internships, lack of communication with other students, and monotonous

lectures, also the lack of a social life, and an abrupt change in routines and habits (Flores et

al., 2021).

On the other hand, 38% had a positive perception, as there were less expenditures

and commutes, better time management, the recording of lectures, more hours of sleep, rest,

18

and leisure, less stress, and better marks. However, students that had a more positive

perception were also able to acknowledge negative aspects such as, poor time management,

worse concentration capacity, the professors’ lack of adaptation to new tools and

methodologies, overwork, and the reduction in learning quality for the same reasons as the

students with a negative perception (Flores et al., 2021).

In the same study, the primary difficulties and negative aspects identified by students

were difficulty in concentrating, poor time management, and difficulties in managing tasks

and assignments (Flores et al., 2021). Also, feeling tired, stressed, and fearful, difficulties in

following the classes from home due to distractions in the environment, technical difficulties

during classes, a lack of support from teachers, and difficulties in answering the teachers’

solicitations (Flores et al., 2021).

The primary facilities and positive aspects identified by students in the study were

having access to the resources they needed for online learning, the possibility of contacting

the professors when having doubts, and the availability of diverse learning materials (Flores

et al., 2021). But also, being able to interact with teachers when necessary, and ease in

following the daily routine of classes (Flores et al., 2021).

Overall, the students' opinion was divided, with contrasting opinions on both negative

and positive aspects, however they were able to identify more negative aspects than positive

aspects. This seems to be due to students’ individual characteristics and learning preferences,

causing several incompatibilities with collaborative platforms. Also, the home learning

environment seems to be impacting students’ level of engagement. These factors were also

identified in the other studies. Furthermore, the abrupt change was a shock to most students,

teachers, and learning institutions, and as the data for this study was collected in 2020, the

first year and the most uncertain year of the COVID-19 pandemic, they probably needed more

time to adjust and adapt to an online learning environment.

From the literature reviewed, we can conclude that there are many points in

accordance, when it comes to difficulties experienced by students, independently of the

country. For online learning students, whether it is in a situation of emergency online learning

or not, there is difficulty in concentrating due to the learning environment at home (Flores et

al., 2021; Lau & Sim, 2020; Liaw, 2008; Serhan, 2020; Soria et al., 2020), a lack of motivation

19

(Alanazi, 2019; Liaw, 2008; Serhan, 2020; Soria et al., 2020), and a lack of human interaction

(Flores et al., 2021; Lau & Sim, 2020; Liaw, 2008; Serhan, 2020; Soria et al., 2020; Y. Zhao et

al., 2005), especially with other students. Plus, greater difficulty in feeling engaged,

connecting it to how interactions work via collaborative platforms (Flores et al., 2021), a

problem highlighted by Webster & Hackley (1997). However, we can sense a difference when

emergency online learning enters at play, as students tend to feel there is a loss in learning

quality (Flores et al., 2021; Serhan, 2020). Therefore, for online learning success there needs

to be a collaborative, communicative, and motivational learning environment, supported by

a software that meets the students' communication, interaction, and learning needs.

2.6 Task-technology Fit theory

In 1995, Goodhue and Thompson proposed a theoretical model that explained how

Information Technology (IT) could have a positive impact on individual performance. That is,

how it could improve the efficiency, effectiveness, and quality of one's performance

(Goodhue & Thompson, 1995). That theoretical model is the Technology-to-Performance

Chain (TPC) model, and it focuses on two constructs: Utilization and Task-Technology Fit (TTF).

TPC predicts that technology is used by individuals as a tool to perform their tasks, which

means tasks are the actions individuals need to carry out, with the support of technology, to

turn inputs into outputs (Goodhue & Thompson, 1995). Therefore, TTF can be defined as the

degree of correspondence between the system’s functionalities or characteristics and the

tasks’ requirements or characteristics (Goodhue & Thompson, 1995). “As tasks become more

demanding or technologies offer less functionality, TTF will decrease” (Goodhue & Thompson,

1995).

According to Goodhue and Thompson (1995), utilization is one's choice to use or not

to use an information system to complete a task or several tasks, and increased utilization will

lead to positive performance impacts. A higher TTF increases the likelihood of utilization and

leads to a positive impact on performance (Goodhue & Thompson, 1995). Sometimes,

utilizing a system can be mandatory. In that case, performance is going to be increasingly

impacted by TTF (Goodhue & Thompson, 1995). Also, organizations can be using a poor

system and that will not increase performance (Goodhue & Thompson, 1995).

20

When information technology is utilized, users can experience effects on performance

originating feedback (Goodhue & Thompson, 1995). From that, users can conclude that the

IT did not meet the expectations or that it surpassed the expected effects on performance,

which will change one’s expected consequences of utilization, influencing future utilization

(Goodhue & Thompson, 1995). Also, users can learn to better utilize the technology,

improving individual technology fit and the overall TTF (Goodhue & Thompson, 1995).

Figure 1 presents a subset of the TPC model, which is the Task-Technology Fit theory.

Although the TPC model was never tested by Goodhue and Thompson, TTF was, in 1995, in

the organizational context. In that study, the authors concluded that TTF and Utilization were

predictors of Individual Performance. However, TTF proved to be the strongest predictor

(Goodhue & Thompson, 1995).

Figure 1: The subset of the Technology-to-Performance Chain model tested by Goodhue and

Thompson (1995).

Because of the COVID-19 pandemic, the use of information technology was something

necessary. Students did not have a say in the choice of the collaborative platform used for

remote classes, so, according to the theory, students’ performance is going to be increasingly

impacted by TTF (Goodhue & Thompson, 1995). Also, if the chosen platform does not perform

the best, student’s individual performance is going to be negatively impacted (Goodhue &

Thompson, 1995). If the platform has features that allow collaboration, sharing, and

communication, the student is going to be able to absorb and share information better, which

in return contributes positively for their learning performance. In other words, if collaborative

platforms’ capabilities provide what students need for performing their tasks, there is going

to be a positive impact on their performance (Goodhue & Thompson, 1995).

21

2.7. TTF and Online Learning

The TTF theory has been used for academic research in a varied set of different

contexts, such as education (Alamri et al., 2020; D’Ambra et al., 2013; Yadegaridehkordi et al.,

2014), mobile banking (Hoehle & Huff, 2012; Oliveira et al., 2014; Tam & Oliveira, 2016b; Zhou

et al., 2010), tourism (D’Ambra & Wilson, 2004; H.-C. Lin et al., 2020; Usoro et al., 2010), and

e-commerce (Chang, 2008; M. Klopping & Mckinney, 2004; Shih & Chen, 2013).

It has also been combined with other models, such as TAM (Dishaw & Strong, 1999;

Pal & Patra, 2020; Shih & Chen, 2013), UTAUT (Oliveira et al., 2014; Zhou et al., 2010), and

Social capital theory (Lu & Yang, 2014).

In the higher education context, TTF has been used in studies for examining university

students' interaction with ICTs and how they can affect students’ performance in online

education. And although technology has been widely used in the educational context, most

studies focus on intention to use, use, adoption, and continuance of use, and there are very

few that approach the impact of technology on students’ performance. For our study, we are

focusing on reviewing the literature directly related to higher education and TTF. Those

studies are going to be presented in this section.

McGill and Hobbs (2008) compared the perceptions of students and instructors on

TTF, user satisfaction, utilization, anticipated consequences of use, social norms, facilitating

conditions, utilization, and performance impacts of the virtual learning environment (VLE)

WebCT. The authors found that for students with a higher TTF, VLE WebCT would have a

greater impact on their performance. Also, students and instructors had the same level of

utilization, but instructors perceived higher levels of facilitating conditions.

McGill and Klobas (2009) studied the role of TTF in the online learning environment,

using the PLS-SEM technique to test the relationships in their model. In their study, the

authors considered both the students' perception of the LMS's impact on performance and

objectively measured performance through student's grades. Facilitating conditions, a

precursor of utilization in the TPC model, had no impact on utilization. TTF had a strong

positive direct influence on perceived performance and a weak positive direct impact on

student's grades. Finally, the authors found that utilization positively influenced students’

perceived performance.

22

Raven et al. (2010) conducted a study on the use of digital video tools for oral

presentations as a substitute for in-class oral presentations, at the university level. The

authors redefined TTF by giving it three dimensions: Task Match, Ease of Use, and Ease of

Learning, and used Usefulness as a mediator of the relationship between Fit and Performance.

They concluded that Fit had a positive direct impact on Usefulness and Usefulness had a

positive direct impact on Performance (Raven et al., 2010).

Yi et al. (2016) investigated college students’ use of smartphones for academic

purposes and how TTF of smartphones can affect students’ performance. The authors

concluded that TTF of smartphones directly influenced students’ performance and indirectly

affected utilization of smartphones through the precursors: attitude toward smartphone use,

social norms, and facilitating conditions.

Harrati et al. (2017) explored academics’ use of online education systems to perform

their tasks in the academic context and discovered that TTF had a greater impact on

performance than on utilization. When analyzing users’ age, gender, academic qualifications

and experience, age was the factor that most influenced utilization and performance.

Wu and Chen (2017) combined the technology acceptance model (TAM) and TTF to

understand continuance intention to use Massive Open Online Courses (MOOCs) and realized

that the greater the fit between the individual, the task, and the technology employed, the

more students would find MOOCs useful and easy to use. Performance was not approached

in this study.

Bere (2018) applied the TTF theory using PLS-SEM, for examining how Mobile Instant

Messaging (MIM) impacted university students’ mobile learning performance, in South Africa.

The results showed TTF impacted Performance (Bere, 2018).

Isaac et al. (2019) conducted a study with 448 students from 9 public universities, in

Yemen, in an online learning course. It is important to note that Yemen ranks last in the

indicator quality of education system, among 138 countries, besides, it is among the world's

lowest GDP per capita countries. The results showed TTF has a direct positive impact on

performance, and compatibility had a direct positive impact on use and user satisfaction

(Isaac et al., 2019).

23

Alanazi et al. (2020), in their investigation, based on the TTF theory, studied what

characteristics of fully online classes account for students’ performance. The strongest

predictor was task value, which is how learners view the task as being important, interesting,

and useful for their educational needs, followed by quality of content, which, also, has a

strong positive influence on task value itself (Alanazi et al., 2020).

Alamri et al. (2020) conducted a study on Social Networking Applications (SNAs) for

collaboration in higher education and found a direct positive path from TTF, Students’

Satisfaction, and Usage to Performance.

Pal and Patra (2020) studied university students’ perception of Video-Based Learning,

in the context of COVID-19, in Science, Engineering, and Business Management faculties

across two big Indian universities. By combining TAM and TTF, the authors realized individual

characteristics, which is defined as prior experience in using similar technology, had a greater

impact on TTF than Technology Characteristics. Performance was not approached in this

study.

From reviewing these studies, we can assess that the TTF theory has been applied in

the higher education context, and there are some attempts in extending the model and in

studying what factors affect students’ performance. However, there is still work to do, and

there needs to be a greater focus on studying what predicts performance.

3. Research Model

Collaborative platforms have been used in organizations, for decades, and more

recently, in education, mostly by students enrolled in distance education courses, but also by

students attending mixed courses, and even traditional courses. Because of the COVID-19

pandemic, the entire student population had to attend classes remotely and experience the

use of collaborative platforms. The very abrupt change, implemented by governmental

organs, generated an unprepared and fast adaptation from educational institutions (Flores,

2020). In addition, the classes for face-to-face courses were not developed thinking of an

online learning environment, that is why collaborative tools were an important asset during

the pandemic. Since they help to emulate face-to-face classes, they helped in creating a more

seamless transition for students and professors. This highlights why online collaborative

24

platforms, now more than ever, need to be studied in relation to students’ individual

performance.

With this study, we are looking to develop a model that explains students’ individual

performance, in an online learning context, with a strong out-of-sample predictive power. We

aim to understand if the collaborative platforms used by students for online learning during

the COVID-19 pandemic fit the tasks they needed to perform and if they can create a positive

impact on students' individual performance. Furthermore, it is important to understand what

other factors can impact performance. To study this matter, the TTF theory developed by

Goodhue and Thompson (1995) will be combined with other constructs found to have a

possible impact on the performance of students attending online classes. The proposed

model is presented in Figure 2.

Figure 2: Research model with the proposed hypothesis.

The research model is composed of 11 variables, of which two are moderator

variables. This model can be read as follows: Task-Technology Fit (TTF), is predicted by Task

Characteristics (TASK) and Technology Characteristics (TECH), and this relationship is affected

by a moderator variable, which is Facilitating Conditions (FC). Utilization (USE) is predicted by

FC, and TTF. However, the relationship between TTF and USE is impacted by Voluntariness

(VLNT), which is moderator variable. Engagement (ENG) is predicted by Environment (ENV),

and Compatibility (COMP) is predicted by Perceived Usefulness (PU). Finally, the dependent

25

variable Individual Performance (PERF), which is the center of our study, is dependent on six

variables PU, COMP, TTF, ENV, ENG, and USE.

Following, we are going to define and explain the relationships we hypothesized in our

research model.

TASK -> TTF & TECH -> TTF

To study what impacts Individual Performance, Goodhue and Thompson (1995),

proposed and tested the TTF Theory. According to that theory, the collaborative platforms’

features and possibilities will affect TTF (Goodhue & Thompson, 1995). Furthermore, the

characteristics and requirements of the tasks performed by students using collaborative

platforms will have an impact on TTF (Goodhue & Thompson, 1995). Therefore, when the task

requirements are supported by the functionalities of collaborative platforms there is a higher

level of TTF (Goodhue & Thompson, 1995).

There are several studies that tested the impact of TASK and TECH on TTF, in the

context of higher education. Raven et al. (2010) tested the effect of TASK and TECH on Fit, in

the context of digital video tools used for oral presentations and found a significant positive

relationship. It is interesting to note that TECH had a greater impact on TTF.

Bere (2018) applied the TTF theory in the context of mobile learning and found very

similar strengths in the positive relationships between TASK and TTF and TECH and TTF, but

TASK had a slightly stronger path.

Wan et al. (2020), studied the factors leading to university students' continued

intention to use Massive Open Online Courses (MOOCs). They found a negative insignificant

relationship between TASK and TTF but, on the other hand, a very strong relationship

between TECH and TTF, in fact it was the strongest path in their model.

Pal and Patra (2020), when studying Video-Based Learning in higher education, tested

the relationship between TECH and TTF and found a positive significant impact.

Considering these findings, we expect TASK and TECH to have a positive impact on TTF.

However, and considering the pandemic situation, there is a reasonable chance TECH is going

to have a greater impact on TTF than TASK. Therefore, we propose the following hypothesis.

H1: Task Characteristics have a positive impact on Task-Technology Fit.

26

H2: Technology Characteristics have a positive impact on Task-Technology Fit.

TASK*FC -> TTF & TECH*FC -> TTF

In our model, FC serves as a moderator variable, influencing the relationships between

TASK and TTF and TECH and TTF. It is a continuous moderator variable, which means it can

affect the direction and strength of the relationship between the exogenous latent variables

(i.e., TASK and TECH) and the endogenous latent variable (i.e., TFF) (Hair et al., 2021). That

means, the relationship between TECH and TTF changes depending on the level of FC, but if

the moderator is not present, the strength of the relationship between TASK and TTF is

constant. The same applies to TECH.

FC is defined by (Triandis, 1979) as the objective factors present in the environment

that allow behavior to occur and make an act easy to do. Thompson et al. (1991), in the

context of IS, developed a scale for measuring FC, and in their scale the items were related to

technical support, however they recognized that other types of FC should have been included.

From that, Venkatesh et al. (2003) developed a comprehensive four item scale for measuring

FC, and in that scale one item was taken from the scale developed by Thompson et al. (1991)

and the other three items were from Perceived Behavioral Control (PBC) (S. Taylor & Todd,

1995). For the context of IS, S. Taylor and Todd (1995) modeled their construct of PBC from

the scale of FC developed by Thompson et al. (1991).

It is important to note that, even though Venkatesh et al. (2003) state that COMP was

used to create the new scale for measuring FC, any of the items developed by Moore and

Benbasat (1991) to measure COMP were included in their new scale. We do not consider FC

and COMP synonyms, and they are going to be studied as different constructs, in our study.

We think COMP should not be ignored or be placed in the same box as FC.

Therefore, we define FC as objective factors that alleviate barriers to use, and

therefore, make using a technology easier (Thompson et al., 1991; Triandis, 1979; Venkatesh

et al., 2003). We consider FC, having the knowledge to use collaborative platforms, having the

necessary resources to use them, fitness with other technologies used by students, and

having available technical support.

27

To our knowledge, FC has not been tested as a moderator of the effect of TASK and

TECH on TTF. However, it has been tested by Gerhart et al. (2015), in the context of e-

textbooks for learning, using the TTF theory, and as a predictor of TTF, and the authors found

a direct, positive, and significant relationship between FC and TTF. This means, as e-textbooks

are easily accessible using laptops, tables, and even mobile phones, students are more likely

to perceive a higher TTF (Gerhart et al., 2015).

The same findings can be easily extended to collaborative platforms. When students

perceive there are many factors making the use of collaboration platforms easier, they are

more likely to feel tasks they need to perform are a good fit for collaborative platforms and

vice-versa. The resources available, such as a computer, and an internet connection with a

large bandwidth, digital literacy, and available technical support can differ for each student.

Therefore, we suggest that when there is a higher level of FC, that is students have the

necessary knowledge, resources, and help, there is going to be an increase in the strength of

the relationship between TASK and TTF and TECH and TTF. This can be illustrated by the

following hypothesis:

H3: The relationship between Technology Characteristics and Task-Technology Fit is positively

moderated by the level of Facilitating Conditions.

H4: The relationship between Task Characteristics and Task-Technology Fit is positively

moderated by the level of Facilitating Conditions.

TTF -> PERF

Goodhue and Thompson (1995), in their study, identified that TTF had a significant

impact on individual performance, accounting for 14% of the variance in PERF.

Gerhart et al. (2015) in testing the impact of e-textbooks on students’ learning

performance found a positive significant impact of TTF on PERF. In a similar context, Jardina

et al. (2021) tested a model for predicting e-textbooks' impact on students' PERF and

confirmed the positive significant impact of TTF on PERF. Bere (2018), in the study of mobile

learning, also confirms this relationship. As well as McGill and Klobas (2009) in the study of

LMSs and Harrati et al. (2017) in the study of online educational systems.

28

Considering these findings, in the context of higher education, it is safe to assume the

existence of a similar relationship in our study. We propose that when there is a higher Fit

between the tasks students need to perform and the collaborative platform they are going to

use to perform those tasks, there is going to be a greater positive impact on PERF. For

example, a student needs to attend classes using collaborative platforms, and if Zoom is

indicated to do so successfully, then there is going to be a greater positive impact on the

student’s performance. Therefore, we propose the following hypothesis.

H5: Task-Technology Fit has a positive impact on Individual Performance.

TTF -> USE

Goodhue and Thompson (1995) found that TTF predicts USE, but it accounts for only

2% of the variation in USE. However, even though the support is little, it is still statistically

significant. Similarly, Dishaw & Strong (1999), when combining TAM and TTF, found a positive

but low impact of TTF on USE.

However, a low impact is not always the case, Gerhart et al. (2015) in testing the

impact of e-textbooks on students’ learning performance found a positive significant impact

of TTF on USE, in fact it was the strongest path in their model.

Considering these findings, we expect students who find there is a fit between TASK

and TECH to be more inclined to use collaborative platforms. Therefore, a higher TTF will

increase USE, and so we propose the following hypothesis

H6: Task-Technology Fit has a positive impact on Utilization.

TTF*VLNT -> USE

In our model, Voluntariness serves as a moderator variable, influencing the

relationships between TTF and USE. It is a continuous moderator variable, which means it can

affect the direction and strength of the relationship between the exogenous latent variable

(i.e., TTF) and the endogenous latent variable (i.e., USE) (Hair et al., 2021). That means the

relationship between TTF and USE changes depending on the level of VLNT, but if the

moderator is not present, the strength of the relationship between TTF and USE is constant.

VLNT of utilization can be defined as the way an individual perceives using a

technology is voluntary, or their own choice (Moore & Benbasat, 1991; Rogers, 1983). It has

29

been studied in relation to intention to use, use, and adoption of an information system, and

as a moderator of the relationship between Subjective Norm and Intention to Use.

Goodhue and Thompson (1995) in their paper about the TPC model, theorized VLNT

should have an impact on TTF and USE. According to the authors when the USE of a system is

mandatory, PERF is going to be increasingly impacted by TTF. However, they did not include

this variable in their model, and so, did not test it.

Agarwal and Prasad (1997) were the first to test and validate if perceived voluntariness

played an actual role in technology acceptance. They tested its influence on current use and

future use intentions of the World Wide Web service available on the Internet, and the

subjects of the study were students enrolled in an MBA program. The results showed that

voluntariness was significant in explaining the current use of an innovation and that it had a

negative impact, but it had no impact on future use intentions. The result suggests that initial

usage of a technology might be influenced by someone obligating an individual to use it, but

individuals will continue to use a technology if they find it useful.

Lee (2006) studied the factors affecting the adoption of Information systems and

tested the moderating role of voluntariness in the relationship between subjective norm and

Intention to use, in the context of e-learning. However, there was no moderating role,

contrary to the findings of Venkatesh and Davis (2000) and Venkatesh et al. (2003).

Widiantoro and Harnadi, (2019) approached the variables affecting the adoption of e-

learning and realized that mandatory users had a negative influence on the variables. This

means, students in a mandatory setting perceive less impacts, when compared to voluntary

users (Widiantoro & Harnadi, 2019). That negative impact was more significant for the

variables Attitude Towards Use, Perceived Ease of Use, Behavioral Intention to Use, and TTF.

Because of the COVID-19 pandemic, the use of IT was something necessary and

students did not have a say in the choice of the collaborative platform used for remote classes.

This means, even if they did not consider there was a fit between the task they had to perform

and the collaborative platform they had to USE, they were still required to use it. Therefore,

the moderating role of VLNT is very important to help explain the relationship between TTF

and USE. Furthermore, measuring VLNT will help us understand the idea of performance

30

being increasingly impacted by TTF with mandatory USE (Goodhue & Thompson, 1995), and

help explain the impact of USE on PERF.

Considering these ideas and the literature reviewed, we expect VLNT, which is

measured by the students’ perception on how much teachers require them to use

collaborative platforms, to have a negative impact. That means, the less voluntary

collaborative platforms’ use is, the more negatively the relationship between TTF and USE is

going to be impacted. Therefore, we suggest the following hypothesis:

H7: Voluntariness has a negative moderating effect on the relationship between TTF and

Utilization

FC -> USE

Goodhue and Thompson (1995) in their TPC model suggested but did not test, FC as a

precursor of USE. However, Venkatesh et al. (2003) and Venkatesh et al. (2012) proposed the

UTAUT, where FC was found to have a significant impact on USE in both organizational and

consumer contexts.

In the context of online education, Khechine et al. (2020) found a significant positive

relationship between FC and USE, in fact, it was the most significant contributor to explaining

use. The same findings are true for Zainol et al. (2017), in a study of mobile learning. Likewise,

Masadeh et al. (2016), in the context of e-learning, found a strong positive relationship

between FC and behavioral intention to use. This means FC can be an important variable for

online learning students.

However, this is not always the case. McGill and Klobas (2009) approached the

variables affecting the adoption of e-learning and did not find FC had an impact on USE.

Similarly, Widiantoro and Harnadi (2019) did not find a relationship between FC and intention

to use. This could be due to there being an indirect effect or direct effect with other related

constructs that were not accounted for.

Considering these findings, we hypothesize that when a student has the necessary and

adequate resources, the knowledge, and technical support for using collaborative platforms,

the student is going to be more inclined to utilize collaborative platforms. Therefore, better

FC leads to increased USE, and so we propose the following hypothesis:

31

H8: Facilitating Conditions have a positive impact on Utilization.

USE -> PERF

Goodhue and Thompson (1995) found that USE had a small impact on predicting PERF.

According to their theory, USE only explained 4% of the variance in PERF.

In the context of education, McGill and Klobas (2009) found a positive relationship

between USE and perceived performance. Harrati et al. (2017) also found a positive

relationship between USE and PERF. Likewise, Alamri et al. (2020) found a significant positive

relationship between Social Networking Sites Usage and PERF. However, in these studies TTF

was found to have more influence on PERF than actual USE.

However, Gerhart et al. (2015) in testing the impact of e-textbooks on students’

learning performance found a positive significant impact of USE on PERF, which was larger

than the impact of TTF on PERF. This could be due to the technology’s nature.

Considering these findings, we hypothesize that with increased USE there is going to

be greater PERF, which can be translated into the following hypothesis:

H9: Utilization has a positive impact on Individual Performance.

ENG -> PERF

To further understand the impact of collaborative platforms on students’

performance, students’ level of engagement in remote classes should be accessed.

There are three types of student engagement: cognitive engagement, which is when

students invest cognitive resources and are willing to make an effort in acquiring the required

knowledge and skills; emotional engagement or presence, which is when students are

motivated and committed and feel like they are part of the class and are fully present; and

behavioral engagement that is, students on-task behavior and conduct (Axelson & Flick, 2010;

Francescucci & Foster, 2013; Northey et al., 2015; Picciano, 2002).

For decades, Engagement has been considered a fundamental factor for measuring

students’ learning outcomes. Francescucci and Rohani (2019), considered engagement one of

the determining factors for comparing online learning courses and F2F courses, and

concluded that online learning students were less engaged but that did not reflect

significantly on students’ performance.

32

Davis & Frederick (2020), in their study about online learning design for best

performance outcomes, noted that diversified multimedia had the ability to reduce cognitive

load and facilitate engagement. In fact, engagement was the strongest predictor of high

student performance, so multimedia should be used strategically to promote that.

Serhan (2020), when measuring students’ attention to class tasks during Zoom

sessions, in comparison to F2F sessions, noted that that was the item with most negative

responses. Students believed Zoom had a negative impact on their engagement, as it did not

promote interaction, with only 13% of students considering that Zoom helped them to

participate in classes.

There have been multiple studies that address student engagement in the context of

online learning, as this is one of the major factors for originating better learning and teaching

outcomes (Axelson & Flick, 2010; Groccia, 2018; Parsons & Taylor, 2011; Robinson &

Hullinger, 2008).

When the student is not engaged, he is not committed to the task and so, his

performance is likely to decrease. Therefore, we hypothesize that with increased ENG, there

is going to be an increase in PERF.

H10: Engagement has a positive impact on Individual Performance.

ENV -> ENG & ENV -> PERF

There are multiple studies that approached how a favorable environment with the

right conditions can impact positively students’ focus, commitment, and attention when

learning, and how well they can perform on academic tasks (Adeniyi et al., 2021; Holley &

Oliver, 2010; Murphy & Yum, 1998; Younas et al., 2021). However, most students, when they

find themselves in an unfavorable environment, can relocate to a better environment, being

the faculty’s library, the work office, or their home space (Murphy & Yum, 1998).

During a crisis, this becomes complicated, as students are confined to their homes,

along with their families, making the environment surrounding the learner even more

challenging. That is why in studies about education and the impact of COVID-19, the home

environment has especially proved to be an important factor (Soria et al., 2020).

33

We define the surrounding environment as the factors that are not controlled by the

student but contribute to the space where the student is going to accomplish academic tasks.

The ENV could have a positive impact on ENG and PERF if the student was provided with a

favorable ENV, with no distractions, a good internet connection, encouraging people, and

little career and family affairs to balance with academic tasks. This means that the more

favorable the surrounding environment is, the more positive its impact on ENG and PER is

going to be. Therefore, we propose the following hypothesis:

H11: The conditions of the Environment surrounding the learner have a positive impact on

Engagement.

H12: The conditions of the Environment surrounding the learner have a positive impact on

Individual Performance.

COMP -> PERF

Rogers (1983), in his book about the five perceived technology characteristics that

affect adoption, divides the definition of COMP into three dimensions. For an innovation to

be compatible it must be consistent with the existing socio-cultural values and beliefs, past

experiences, and ideas, and needs for innovations of potential adopters (Rogers, 1983). This

means COMP is connected to the individual's characteristics and it conveys the idea of the

technology fitting with the individual using it.

Incompatibility with strongly held values, especially cultural values, means the

innovation is going to conflict with the adopter and it will be adopted very slowly or there will

be reluctance to make the change (Rogers, 1983). In our context, this could be viewed as

collaborative platforms conflicting with the inherent aspects of a course typology or the

traditional methodology used for teaching a course (Moore & Benbasat, 1991).

The adopter of an innovation has previously adopted ideas, which are used for

assessing new ideas. They work as a familiar standard and when the innovation is close to

that standard, uncertainty decreases. When an innovation is totally compatible with the

adopter's existing practices, it represents no change and therefore there is no innovation from

the perception of the potential adopter (Rogers, 1983). This can be adapted to our context as

collaborative platforms being compatible with the way students like to work and learn

(Cheung & Vogel, 2013; Moore & Benbasat, 1991).

34

For an innovation to be compatible it must meet the needs of the adopter. However,

the adopters might only realize they need an innovation when they become aware of it. This

means that when the needs are fulfilled by the innovation, there is a faster rate of adoption

(Rogers, 1983). In our context, this can be translated as collaborative platforms’

functionalities serving the students’ needs, in terms of their education (Raven et al., 2010)

and the students’ current situation (Moore & Benbasat, 1991).

The construct COMP is related to the characteristics of an individual being compatible

with a technology. It has been linked to use (Cheung & Vogel, 2013) but should also be tested

for having an influence on PERF, as TTF, which is related to the TASK being compatible with

TECH, has already been. To our understanding, the COMP of an individual with the

functionalities of collaborative platforms should benefit students’ PERF, as when a technology

is adequate for a student’s current situation, course, needs, and preferred way of learning,

the better should a student’s PERF be. Therefore, we propose the following hypothesis.

H13: Compatibility has a positive impact on Individual Performance.

PU -> PERF

Perceived Usefulness can be defined as the degree a user perceives using a technology

would enhance his or her job performance (F. D. Davis, 1989), and it was first theorized and

tested in the context of information systems as being part of the TAM (Davis, 1989).

In the context of online learning, PU has been studied to understand systems

adoption, and it has been linked to attitude towards use and behavioral intention to use (M.

K. O. Lee et al., 2005; Y. Lee, 2006; Widiantoro & Harnadi, 2019).

Raven et al. (2010) tested the effect of usefulness on performance and found a

significant positive relationship where usefulness explained 30% of the variance in

performance. Similarly, Alanazi et al. (2020) studied the impact of task value, which is defined

as the usefulness of a task, on performance and found a significant and positive relationship.

We propose that when students find collaborative platforms useful and advantageous,

they will feel more motivated to use them purposefully for learning and performing tasks

(Alanazi et al., 2020). Also, when collaborative platforms allow students to complete academic

tasks quickly, allow them to be more productive, ease the performance of tasks, and are

35

useful and advantageous for learning, there is probably going to be a better outcome when it

comes to students’ performance and learning (Chan et al., 2016; Moore & Benbasat, 1991;

Raven et al., 2010). Therefore, we propose the following hypothesis.

H14: Perceived Usefulness has a positive impact on Individual Performance.

PU -> COMP

Alamri et al. (2020), conducted a study on Social Networking Applications (SNAs) in

the context of higher education and found a direct positive and significant relationship

between PU and COMP.

We propose that, when students perceive collaborative platforms as useful to

enhance their academic performance, they will perceive them as more compatible with their

needs and course. Therefore, we hypothesize that:

H15: Perceived Usefulness has a positive impact on Compatibility.

Mediating role of engagement and compatibility

A simple mediation model represents the indirect effect of one intervening variable

(Hayes, 2009). That is, there is one construct in the middle of a relationship between two

other constructs (Hair et al., 2021). In our model, we are focusing on the mediating effects of

Engagement and Compatibility.

Figure 3 - Engagement mediation model

36

In the first simple mediation model (Figure 3), ENV predicts ENG and PERF is predicted

by both ENV and ENG. c’ represents the path that quantifies the direct effect of ENV on PERF.

The product of a and b quantifies the indirect effect of ENV on PERF through ENG.

Murphy and Yum (1998), when studying Hong Kong distance learners found a

connection between students’ ENV and their ENG. For example, students with a more

challenging environment, such as having visitors at home, their family not giving them a quiet

environment in times of exams and assignments, and balancing family time, would make ENG,

commitment, and focus harder, which would in turn make the completion of tasks more

difficult. Also, on studies about online learning during the COVID-19 pandemic, this seems to

be a recurring theme (Flores et al., 2021). That is why we expect a direct impact of the ENV

on ENG and PERF, but also an important indirect effect through ENG on PERF.

Therefore, in our model, we propose that the Environment’s total effect on Individual

Performance is due to the sum of a direct effect and an indirect effect, and so, we suggest

hypothesis 16.

H16: Engagement mediates the relationship between Environment and Individual

Performance.

Figure 4 - Compatibility mediation model

In the second simple mediation model (Figure 4), PU predicts COMP and PERF is

predicted by both COMP and PU. c’ represents the path that quantifies the direct effect of PU

on PERF. The product of a and b quantifies the indirect effect of PU on PERF through COMP.

37

Alamri et al. (2020) were the first to test the mediating role of COMP. In their model,

COMP mediates the relationships among user satisfaction, actual use, and academic PERF. In

our simpler model, we propose that Perceived Usefulness’ total effect on PERF is due to the

sum of a direct effect and an indirect effect through COMP on PERF. Therefore, we suggest

hypothesis 17.

H17: Compatibility mediates the relationship between Perceived Usefulness and Individual

Performance.

4. Research Methods

The initial stage of a scientific study is the identification of an important research

problem that is situated, grounded, and diagnosed in the real world, which will lead to the

research question that is going to be investigated so we can better understand the problem

and reach a valuable answer (Rai, 2017; Van De Ven, 2007; Venkatesh et al., 2016).

The research problem of this sequential mixed methods study is collaborative

platforms’ impact on university students’ individual performance. Our purpose is to discover

what factors affect Portuguese students, using collaborative platforms for online learning, in

terms of their individual performance.

In extending a theoretical model, by adding new constructs, it is important to have in

consideration the context being studied (Venkatesh et al., 2012). New contexts, such as a new

technology, a new user population and a new environment or setting, can result in important

changes to theories (Venkatesh et al., 2012).

We propose the extension of the TTF theory by applying it to a new context and by

adding new constructs. That is why the first step of our investigation was to analyze and

understand the context of the study not just in terms of the literature already developed, but

also, in terms of the characteristics of the population and the environment they were living

in. For that, we recurred to government reports and announcements, UNESCO reports,

websites, and other reports.

To find answers to the research questions and hypothesis, which were grounded on

the purpose and context of our investigation, we are adopting a mixed methods research

approach.

38

We adopted an integrative approach to the literature review, so for our integrative

review we assessed, critiqued, and synthesized the literature on the topic of information and

communication technology, in the context of online learning, for higher education, in order

to develop a new theoretical model that focuses on student’s individual performance (Snyder,

2019). The literature review constructed our knowledge base, helped us understand the

environment of our study, and thus provided a direction for the research questions and

hypothesis of the study, and it is going to help us corroborate and support our findings

(Creswell, 2009).

To search for relevant articles, we only considered peer reviewed articles. We used

the Educational Resources Information Center (ERIC), an online digital library for research and

information on the topic of education, Google Scholar, and Semantic Scholar.

When using ERIC's thesaurus to look for topics and keywords, we understood that

collaborative technology, collaborative platforms, and collaborative tools were terms with no

results. Collaboration appears, only attached to the words teaching and learning. So, we

decided to look for terms related to the possibilities and functionalities of collaborative

platforms, and we found the following to be relevant terms: asynchronous communication,

synchronous communication, and videoconferencing. Nevertheless, we still searched the

online library for collaborative platforms, tolls, and technology, and some relevant results

were found. After creating a list with all the major terms and important related terms, we

filtered our research for peer reviewed articles, and education level at higher education.

Google Scholar and Semantic Scholar were used to identify the leading literature and

search for relevant literature within the citations of those leading articles.

4.1. Research Questions

From the research problem, the context, and knowledge base, came the research

questions that are going to be investigated, so we can better understand the problem and

reach a resolution (Van De Ven, 2007). As we are following a mixed methods approach, we

propose 2 research questions. That is, 1 mixed methods research question and 1 major

qualitative research question. Plus, 17 hypothesis for our quantitative study.

As Venkatesh et al. (2016) state, mixed-methods research questions are not formed

the same way quantitative and qualitative research questions are. In the context of mixed

39

methods research and by following Creswell's (2009) guidelines, we are including separate

questions for the quantitative and qualitative research strands, plus a mixed methods

research question that integrates both strands of the research. "This highlights the

importance of both the qualitative and quantitative phases of the study as well as their

combined strength (Creswell, 2009)." Therefore, our mixed methods research question is:

To what extent and in what ways do qualitative interviews with university students

serve to contribute to a more comprehensive and detailed understanding of the

predicting relationship between TTF, ENG, ENV, PU, COMP, USE, and students’

PERF, via an integrative mixed methods discussion?

In a qualitative study, the research question includes one to two central broader

questions and optionally five to seven sub-questions that will help to explore the central

problem in greater detail (Creswell, 2009). A qualitative central question should begin with a

what or a how, use an open-ended exploratory verb, focus on a single phenomenon, and

specify the participants and the research site (Creswell, 2009). Our main qualitative research

question is:

How do Portuguese university students, who attended online classes, during the

COVID-19 pandemic, in 2020 and 2021, perceive collaborative platforms’ impact on

their online learning and individual performance?

And our sub questions are:

How do these university students perceive TTF, USE, ENG, ENV, PU, and COMP,

impact their PERF?

What are the students’ perceptions on the advantages and disadvantages of

collaborative platforms?

How do students feel about the level of VLNT of USE?

What are the students’ preferred education method in a near future scenario?

For quantitative studies, there are quantitative research questions or hypotheses, and

the researcher should only opt for one of them to avoid redundancy (Creswell, 2009). As

suggested by Creswell (2009) our hypotheses are specific and narrow, based on the

40

relationship between variables. In our study we used directional hypotheses, where a

prediction about the expected outcomes is made, based on prior studies and literature

(Creswell, 2009). Those hypotheses have been presented in our Research Model.

4.2. Mixed Methods Design

In our mixed methods study, we opted for a sequential approach, where the

quantitative or qualitative study is done after the other. Being so, first we are going to conduct

a quantitative study, since our study is based on existing dimensions with strong theoretical

foundations and evidence, even though applied to a new context (Venkatesh et. al., 2013).

Second, based on the findings from our quantitative study, we are going to conduct a

qualitative study. This brings us to our first purpose for using a mixed methods approach,

which is the developmental purpose, explained in the next paragraph.

Venkatesh et al. (2013) identify seven purposes for using a mixed methods approach,

which are complementarity, completeness, developmental, expansion,

corroboration/confirmation or triangulation, compensation, and diversity. The authors argue

that the mixed methods approach should serve at least one of those purposes or else it is not

justifiable to use this type of approach instead of using a single method approach. Therefore,

we justify our use of mixed methods with four of those purposes. As it was mentioned before,

we conducted the quantitative study first and the qualitative study after, adopting a

sequential research approach. This goes in line with the developmental purpose, which states

one study is developed from the other’s results and, as our quantitative study provided the

questions to be answered and tested in the qualitative study, we can verify that. We used the

qualitative data with the purpose of gaining complementary insights on the findings from our

quantitative study. Also, we wanted to ensure the completeness of our investigation, so we

used our qualitative data results and findings to paint the complete picture. Lastly, we

intended to corroborate the results obtained, so we could add credibility to our inferences

and support the findings that go against the literature, therefore we seek convergence of

results by having the qualitative results corroborate the quantitative results.

We are going to adopt a Single Paradigm Stance, as we believe using multiple

paradigms does not contribute to a greater understanding of the phenomenon under study.

Pragmatism is the paradigm we adopt as the epistemological foundation for our mixed

41

methods research. According to pragmatism both qualitative results and quantitative results

can be used for corroborating the findings from the other study (Rossman & Wilson, 1985).

In our case, since we are doing a sequential mixed methods study, our qualitative results will

provide convergence in findings. Also, pragmatism allows for the findings from our qualitative

study to be used to interpret and elaborate with more detail the findings from the

quantitative study (Rossman & Wilson, 1985).

As Venkatesh et. al. (2013) noted, in their review of IS research that employed a mixed

methods approach, researchers tend to focus in more detail on one of the methods and the

findings are kept separate from each other, forgetting the most critical and essential aspect

of mixed methods research: meta-inferences. In our study we are going to develop meta-

inferences, so we can provide a rich analysis, where both quantitative and qualitative data

are rigorously analyzed, and where the findings from both studies are integrated for

producing meaningful inferences.

We generated our research model and hypothesis deductively, through the literature

review. That is, from the knowledge and theories that already existed, we advance our own

theory (Creswell, 2009). We collected data through surveys to test our proposed model,

reflected on its confirmation and disconfirmation in the results, and made sure we followed

patterns of validity, reliability, and quality in that process.

For our qualitative research we adopted an inductive logic. That is, through interviews,

we gathered information for our qualitative study. The interviews’ recording allowed us to

analyze our data and organize it by themes. From that we were able to look for patterns and

make generalizations (Creswell, 2009). Finally, by bridging our qualitative findings with the

findings from our previous quantitative study, we were able to form meta-inferences

(Venkatesh et al., 2013; Creswell, 2009)

For our exploratory investigation, we opted for a Mixed Methods design with more

than one stage. Those stages are the conceptualization stage, where the research problem

and theoretical model were established; the experimental stage, which is divided in two

stages, which are the methodological stage, where we planned and designed our study, and

the analytical stage, which includes the data collection and analysis (Teddlie & Tashakkori,

2006). These stages were applied to each strand of our study in that order. We divided our

42

research in two separate strands, one quantitative strand and one qualitative strand, that

occurred chronologically, in that order. In our sequential mixed design, the findings from the

quantitative stage led to the formulation of research questions for the next qualitative strand.

Each strand has equal importance, so we are using an equivalent status design. The last stage

of our investigation is the inferential stage, where the results from both studies are combined

to form meta-inferences and a theory emerges (Teddlie & Tashakkori, 2006).

Figure 5 - Methodological model for our mixed methods approach

Figure 5 illustrates and resumes the methodology employed throughout this study. In

this section we presented the research questions, the purposes of our mixed methods

approach, and explained the flow of our investigation. The design for both quantitative and

qualitative approaches is going to be explained in detail in their own sections, along with the

methods to collect and analyze the quantitative and qualitative data.

43

5. Quantitative Approach Design: Data Collection Methods

5.1. Surveys

To collect primary data, specifically for this research, we used a web-based survey and

administered it online. We used Google Forms for designing our survey because it allows for

the creation of a shareable link, and it automatically produces the data output. We created 4

shareable links, one for each surveyed university. The survey was cross-sectional, that is

collected at one point in time (Creswell, 2009), since the data was collected between the 30th

of March 2021 and the 4th of May 2021.

To ensure there would be no ethical issues in the process of collecting data, the

purpose and title of the study were disclosed to the participants, as well as the identification

of the researchers and institutions they belong to. It was explained why the participants were

selected, a guarantee of their confidentiality was given, and, also, contacts were made

available in case there were questions from the participants (Creswell, 2009).

5.2. Sample

This study was conducted in Portugal and targeted students attending courses at

different Lisbon universities, to ensure robust data and better results, as we wanted our

sample to reflect as best as it could the similarities and differences found in the population

(Hair et al., 2016). The universities included in our study are the Faculty of Medicine (FMUL)

from the University of Lisbon, the Faculty of Human Kinetics (FMH) from the University of

Lisbon, the NOVA School of Science and Technology (FCT) from the NOVA University of Lisbon,

and the NOVA School of Social Sciences and Humanities (FCSH) from the NOVA University of

Lisbon. For our quantitative study we used purposive random sampling, as the sample was

randomly selected from a small number of units, represented by the universities we chose, in

a much larger target population, that is students attending Lisbon universities (Venkatesh et

al., 2016).

5.3. Sample Characteristics

Table 1 - Sample characteristics

Distribution (n=295)

44

Gender Attending Degree

Male 150 51% Bachelor 198 67%

Female 145 49% Post-Graduation 2 1%

Master 91 31%

Number of people living with the student Doctorate 4 1%

0 11 4%

1 27 9% Online Classes before the COVID-19 pandemic

2 69 23% No 273 93%

3 124 42% Yes 22 7%

4 41 14%

5 20 7% Has a Personal Computer

6 3 1% Yes 284 96%

No 11 4%

Age

18-21 205 69% Student-worker

22-25 54 18% No 245 83%

>25 36 12% Yes 50 17%

Platforms Used Universities

Zoom 293 99% FCSH 95 32%

Google Tools 88 30% FCT 120 41%

Microsoft Teams 35 12% FMH 59 20%

FMUL 21 7%

Our quantitative study had 295 participants with a very similar distribution of males

and females. Most of the surveyed students, at the time of the survey, were attending a

bachelor’s degree (67%), about one third were taking their master’s degree, and 83% were

full time students. 63% of the surveyed students were between the ages of 18 and 21 and

65% were living with 2 to 3 other people. More than 90% of the students had never attended

online classes before the COVID-19 pandemic, had their own computer, or used Zoom for

academic purposes, during the COVID-19 pandemic.

To reinforce the quality of our data, it is interesting to note that, in a study about

remote learning, in Portugal, for higher education, during the COVID-19 pandemic, when

analyzing the sample, about 97% of the students had a personal computer, and only about

6% had previous experience with online learning (Flores et al., 2021).

45

5.4. Data collection and measurements

Table 2- Items used to measure the constructs in our research model with its corresponding

citations

Constructs Statements References

Technology Characteristics

Collaborative platforms provide real-time communication.

(Zhou et al., 2010)

Technology Characteristics

Collaborative platforms had no frequent unexpected problems.

(Raven et al., 2010)

Technology Characteristics

I would give an excellent rating to the overall quality of collaborative platforms.

(Raven et al., 2010)

Technology Characteristics

Overall, collaborative platforms have useful functionalities such as instant messaging, video recording, screen sharing, etc.

(Lu & Yang, 2014)

Task Characteristics

Frequently, I had to collaborate with others in my academic tasks

(Raven et al., 2010)

Task Characteristics

To perform my academic tasks, I need to communicate with others anytime and anywhere.

(Zhou et al., 2010)

Task Characteristics

My academic tasks required frequent coordination with the efforts of others.

(Raven et al., 2010)

TTF Collaborative platforms provide a suitable set of functions for my academic tasks.

(Lu & Yang, 2014)

TTF Collaborative platforms have enough functionalities to help me perform my academic tasks.

(Lu & Yang, 2014)

TTF I believe that using collaborative platforms for online learning can meet all aspects of my learning requirements.

(Pal & Patra, 2020)

TTF I believe that using collaborative platforms is appropriate and necessary in an online learning environment

(Pal & Patra, 2020)

Facilitating conditions

I have the necessary resources to use collaborative platforms.

(Venkatesh et al., 2012)

Facilitating conditions

I have the necessary knowledge to use collaborative platforms.

(Venkatesh et al., 2012)

Facilitating conditions

I can get help from others when I have difficulties using collaborative platforms.

(Venkatesh et al., 2012)

Facilitating conditions

Collaborative platforms are compatible with other technologies I use.

(Venkatesh et al., 2012)

Utilization I have already used collaborative platforms for communicating matters related to projects.

(Cheung & Vogel, 2013)

Utilization I have already written and read messages or comments on collaborative platforms.

(Cheung & Vogel, 2013)

46

Utilization I have already used collaborative platforms to attend classes.

(Pal & Patra, 2020)

Utilization Currently, I frequently use collaborative platforms

(Hou, 2012; Thompson et al., 1991)

Voluntariness My teachers require me to use collaborative platforms. (Y. Lee, 2006)

Compatibility Collaborative platforms are compatible with all aspects of my course.

(Moore & Benbasat, 1991)

Compatibility Using collaborative platforms is compatible with my current situation.

(Moore & Benbasat, 1991)

Compatibility I think using a collaborative platform fits well with the way I like to work

(Moore & Benbasat, 1991)

Compatibility The functionalities of collaborative platforms serve my needs perfectly.

(Raven et al., 2010)

Perceived Usefulness

Using Collaborative platforms enables me to accomplish my academic tasks quickly

(Moore & Benbasat, 1991)

Perceived Usefulness

Using collaborative platforms enhances my productivity

(Moore & Benbasat, 1991)

Perceived Usefulness

Using collaborative platforms makes accomplishing my academic tasks easier.

(Moore & Benbasat, 1991; Raven, et al., 2010)

Perceived Usefulness

I believe collaborative platforms are useful learning tools (Chan et al., 2016)

Perceived Usefulness

Overall, I find it advantageous to use collaborative platforms for my academic tasks.

(Moore & Benbasat, 1991; Raven et al., 2010)

Individual Performance

Collaborative platforms have a large positive impact on my performance as a student.

(McGill & Hobbs, 2008; McGill & Klobas, 2009)

Individual Performance

I learn better with collaborative platforms than without them.

(McGill & Hobbs, 2008; McGill & Klobas, 2009)

Individual Performance

With the help of collaborative platforms, I have gained a clear understanding of the subjects taught in class

(W.-S. Lin, 2012)

Individual Performance

With the help of collaborative platforms, I managed to achieve the learning goals asserted for my course

(Lin, 2012)

Individual Performance

Collaborative platforms positively contributed to my grades.

(Saragih et al., 2021)

Engagement I am able to consistently pay attention during online classes.

(Sun & Rueda, 2012)

Engagement I am interested in the tasks done during online classes.

(Sun & Rueda, 2012)

Engagement If I do not understand what was taught online, I watch back the recorded session

(Sun & Rueda, 2012)

Environment At home, I have no distractions that hinder my focus and commitment to academic tasks

(Adeniyi et al., 2021)

47

Environment Combining my career, house chores, and academic tasks have made me academically more effective

(Adeniyi et al., 2021)

Environment The internet connection did not inhibit my online learning

(Adeniyi et al., 2021)

Environment The people I live with encourage me in my online learning

(Younas et al., 2021)

Environment The people I live with provide me with a favorable environment to accomplish my academic tasks.

(Younas et al., 2021)

All the variables in our research model represent unobservable traits indirectly

measured by multiple items. Those items are each directly measured by one statement.

The statements used to measure the constructs in our research model, were adapted

from established measurements, published in prior research studies, as can be verified by

Table 2. Those statements had to be evaluated by the survey’s respondents using a ranging

scale labeled from 1 (i.e., Strongly Disagree) to 7 (i.e., Strongly Agree).

All the survey respondents were Portuguese native speakers, so all the survey items

had to be accurately translated from English to Portuguese. To ensure accuracy and quality in

the translation, back translations were performed (Brislin, 1970).

The first question in our questionnaire assessed the collaborative platforms used by

students during the COVID-19 pandemic. Therefore, a list with the most used collaborative

platforms during the pandemic was developed, and each student had to mark the platforms

they used. In case the list did not contain all the platforms used, students were asked to write

the missing ones. The list contained the following platforms: Zoom, Google Tools, and

Microsoft Teams.

Voluntariness was measured through one indicator inspired by Lee’s (2006) one item

scale, which was originally adapted from the scales of Moore and Benbasat (1991) and

Venkatesh and Davis (2000) to fit the e-learning context.

Compatibility was adopted from Moore and Benbasat’s (1991) scale. However, we

added one item from the scale of Raven et al. (2010) for the construct Task Match, as it goes

in line with Rogers’ (1983) 3-part compatibility definition, which is the base of our definition,

and it accounts for the technology serving the needs of an individual.

48

Each item in engagement accounts for the three types of engagement enclosed in

engagement’s definition. Being so, “I am able to consistently pay attention during online

classes” represents Behavioral Engagement, “I am interested in the tasks done during online

classes” represents Emotional Engagement, and “If I do not understand what was taught

online, I watch back the recorded session” represents Cognitive Engagement.

Lastly, a demographic questionnaire was placed to characterize our sample and

account for social differences that could have an impact on the constructs being analyzed.

The collected sample data was used to test for the rejection or nonrejection of the

proposed hypotheses, which led to our results.

6. Quantitative Approach Design: Data Analysis

The objective of the quantitative analysis was to test the relationships in our

theoretical model through the PLS-SEM method with data collected from a survey employed

to 295 university students that had to attend online classes during the COVID-19 pandemic.

Being TTF, an established Management Information Systems (MIS) theory for assessing

Individual Performance, we wanted to test it in the context of online learning and extend it

with other variables that seemed to have an impact on students’ performance. We aim to

understand if task characteristics and technology characteristics would have a similar or

contrasting impact on TTF, and if utilization and TTF would have a significant impact on

individual performance, and which would be more significant. In addition, we wanted to

explore facilitating conditions as a moderator of both the impacts of task characteristics and

technology characteristics on TTF. As well as voluntariness as a moderator of the relationship

between TTF and utilization.

Furthermore, the extension of the TTF theory intended to find other relevant

constructs with the ability to empower the prediction of individual performance. Therefore,

we added to the model perceived usefulness, compatibility, environment and engagement as

direct predictors of students’ individual performance. We also wanted to test the mediating

roles of Engagement and Compatibility.

We followed the minimal sample size guidelines, as suggested by Hair et al. (2016), so

we would have an adequate statistical power and avoid committing a type II error, since an

insufficient sample size can hide an effect that exists, negatively affect the robustness of the

49

results, and the model generalization. To estimate our minimal sample size we used the

Minimum R-squared method, which was first presented by Hair et al. (2014) in the first edition

of their book on PLS-SEM (Hair et al., 2016; Kock & Hadaya, 2018). Our model has 6 arrows

pointing at Individual Performance and we are using a significance level of 5%. For a statistical

power of 80%, and to detect a minimum 𝑅2 of 0.10, the worst-case scenario, we would need

a sample with at least 130 participants. We also used the program G*Power to calculate the

a priori required sample size, for a statistical power of 95%, with a significance level of 0.05

and to detect an effect size of 0.10, it would be necessary 132 survey responses. Therefore,

in our data collection, for our quantitative studies, we used the value 132 as a baseline when

collecting survey responses.

We pilot tested our questionnaire with 52 participants, to adjust items in the survey,

if necessary. According to our pilot test, there was no need to alter or remove any items.

Before testing our model with the data obtained from the survey responses, we

examined the data to prepare for the application of the PLS-SEM method. There were no

missing values, suspicious survey response patterns, or outliers.

Structural Equation Modeling (SEM) is a second-generation technique for multivariate

data analysis, and it started to be used in social sciences in the 1970s (Hair et al., 2011, 2021).

It models the dependent and independent variables, which are measured indirectly by more

than one indicator, and it estimates the complex relationships among these variables (Hair et

al., 2021). Partial Least Squares (PLS) is a type of SEM method that uses a causal-predictive

approach, by providing causal explanations of the relationships estimated in the structural

statistical model, maximizes the explained variance in the dependent variables, and evaluates

the data quality through the measurement model (Hair et al., 2011, 2019).

We adopted the PLS-SEM method for analyzing our quantitative data, as it is indicated

for exploratory research with complex predictive theoretical models (Hair et al., 2011, 2019),

which is our case, since we are extending an established theory and increasing its complexity.

Furthermore, there was little knowledge on how the added variables, such as compatibility

and perceived usefulness, were related to performance (Hair et al., 2019, 2021).

SmartPLS 3 (version 3.3.3) is the analytical tool that is going to help us test the

relationships between the multiple variables in our model. It combines Principal Component

50

Analysis (PCA), which is responsible for telling us if the indicators we used for measuring our

constructs are valid and reliable, with ordinary least squares regression analysis, which is

applied to test the relationships between the constructs, telling us if they are positive or

negative and significant or non-significant (Hair et al., 2019).

For our resampling method, we used Bootstrapping, a non-parametric procedure that

creates subsamples from random observations in the original set of data (Hair et al., 2021). In

our case, we chose to create 10.000 random subsamples, which is a large enough number to

ensure highly similar results for several bootstrap runs, thereby providing a stable basis for

drawing inferences (Hair et al., 2021; Sarstedt et al., 2020). For our confidence interval

method, we used the Bias-Corrected and Accelerated (BCa) Bootstrap method, as it is

recommended by Hair et al. (2021) and Sarstedt et al. (2020). For our two-tailed test we used

a 5% significance level (i.e. p-value > 0.05), meaning there is a 5% chance or less of the

relationship between the variables being mistakenly supported (Goodhue et al., 2006), since

it is the most commonly used in studies (Hair et al., 2016, 2021). We set the maximum number

of iterations to 300, as this number should be sufficiently large to ensure the algorithm

stopped due to the stop criterion and not because the maximum number of iterations was

reached (Dakduk et al., 2019; Hair et al., 2013). The stop criterion value was set to 10−7so

there is a minimal change (<0.0000001) in the outer weights between two iterations (Dakduk

et al., 2019; Hair et al., 2013, 2021). We applied the path weighting scheme (Hair et al., 2011,

2016) and used +1 as the initial value for the outer weights of each indicator (Hair et al., 2013).

7. Quantitative Results

This chapter presents the detailed results of the test of our SEM using the PLS

method. First, we are going to review the Measurement Model, which assesses the reliability

and validity of the constructs. Validity will tell us if our data is plausible, credible, and

trustworthy, and thus can be defended when challenged (Venkatesh et al., 2013). Second, we

are going to analyze the Structural Model, which helps assess the relationship between

variables and test our hypothesis. After careful evaluation of both models, the results are

going to be interpreted so we can draw our first findings.

51

7.1. Measurement Model

Our model has ten constructs measured by observable multiple items, and one

construct, which is voluntariness, measured by one single item, with each item being related

to a survey question. All the constructs in our model have a reflective measurement model.

The reflective measurement model is the outer model and it presents the unidirectional

predictive relationships between a construct and multiple indicators, with an outwards

direction, starting from each construct and finishing on its associated observed indicators

(Hair et al., 2011). An indicator’s outer weight reflects how important that indicator is for the

construct, since indicators with a higher outer loading contribute more to the construct, and

indicators with a lower weight have a high degree of measurement error, and thus

contributing less (Hair et al., 2021).

In evaluating our reflective measurement model, we are going to assess internal

consistency, through Cronbach’s Alpha and Composite Reliability; convergent validity,

through indicator reliability and Average Variance Extracted (AVE); and discriminant validity,

through cross-loadings, Fornell-Larcker criterion, and heterotrait-monotrait (HTMT) ratio of

correlations.

To do so, we are using the SmartPLS 3 software, and running the PLS Algorithm with a

path weighting scheme and a maximum of 300 iterations. And the stop criterion value was

set to 10−7 (Dakduk et al., 2019; Hair et al., 2013, 2021). The algorithm converged at seven

iterations.

We are going to start by evaluating the constructs' internal consistency or reliability.

Engagement showed problems, since it has a minimally acceptable value for the Cronbach’s

Alpha, that is between 0.60 and 0.70 (Dakduk et al., 2019), even though Composite Reliability

had satisfactory values, that is between 0.70 and 0.95 (Hair et al., 2016). This happens because

the Cronbach’s Alpha is a more conservative measure that tends to underestimate the

internal consistency reliability, and Composite Reliability tends to overestimate it (Hair et al.,

2016). Therefore, the true internal consistency reliability should be in between those two

measures (Hair et al., 2016).

Considering that, Engagement has a Cronbach’s Alpha of 0.670 and a Composite

Reliability of 0.819. This Cronbach’s Alpha’s value means there is an indicator measuring its

52

respective latent variable that is not equally reliable or closely related to the other indicators

measuring the same latent variable (Dakduk et al., 2019; Hair et al., 2016).

When the reflective indicators are consistent and equally reliable, removing an

indicator should not alter the construct consistency, because the remaining indicators would

still adequately represent the construct (Bollen & Lennox, 1991; Jarvis et al., 2003). On the

other hand, if there is an unreliable and inconsistent indicator, removing it should benefit the

construct’s consistency.

To verify that, in a reflective measurement model, an indicator should be reliable if

the outer loading is higher or equal to 0.70 (≥0.70), however in exploratory studies outer

loadings of 0.40 or higher (≥0.40) can be accepted (Hair et al., 2013, 2016; Sarstedt et al.,

2014). Even though not all indicators are equally reliable, different outer loadings means that

the indicators are measuring different phenomenon, and when combined, they can still be

consistent, as a group (Dakduk et al., 2019; Hair et al., 2016).

Being so, to decide if we are going to exclude an indicator, we are going to look at the

outer loadings. There was one indicator from Engagement that clearly showed a lower outer

loading, that is 0.431, but it can still be included in the model if removing it will not increase

the Cronbach’s Alpha above 0.70 and negatively impacts validity (Hair et al., 2016). After

removing the third indicator from Engagement, all the criteria for evaluating the

measurement model’s reliability and validity improved, as can be seen in Table 3. It seems

the item measuring cognitive engagement was not consistent, and therefore it is not reliable.

Table 3 – Engagement’s indicators before and after removing the indicator ENG3

Indicators Outer Loadings

Outer Loadings

Cronbach's Alpha

Composite Reliability

Average Variance Extracted (AVE)

ENG1 .911 .939 .67 -> .858 .819 -> .933 .623 -> .875

ENG2 .924 .932

ENG3 .431 removed

53

It is important to note that a Composite Reliability above 0.95 is not good, because it

means the indicators are measuring the same phenomenon (Hair et al., 2021). This could

hinder the validity of the construct’s measure, as it could mean there are redundant and

repetitive items in the outer model. ENG, PU, COMP, and PERF, all have Composite

Reliabilities above 0.90 but they still are below 0.95. Because we are considering a less

conservative threshold (i.e., 0.95), there is no need to remove indicators.

The outer loadings are the coefficients that quantify the relationship between the

construct and the indicator, and determine how well the indicators will measure the construct

and thus successfully converge (Hair et al., 2011). Therefore, to determine convergent

validity, we are going to assess if the indicators measuring a construct have the ability to give

at least 50% of their variance to the construct they are defining. For that, an indicator’s outer

loading should be above 0.708, so they can converge for sharing a high proportion of variance,

and lead to an Average Variance Extracted (AVE) with a value above 0.50, which is the

minimum threshold (Hair et al., 2016, 2021). But, if the outer loadings are below 0.708, and

as long as the outer loading’s value is not lower than 0.40, indicators should only be removed

if doing so increases the AVE (Hair et al., 2016). Also, removing an indicator just because of a

lower value in the outer loading could lead to problems in the construct's validity (Hair et al.,

2016).

The Average Variance Extracted for all the constructs is above 0.50, which means that

on average the construct explains more than half of the variance of its indicators. However,

there are three constructs with indicators with outer loadings below 0.708 but none below

0.40. Before removing them, there should be a careful evaluation to ensure the validity of the

construct is not damaged and if there is a raise in the Average Variance Extracted.

The first construct under evaluation is Task-Technology Fit (TTF), its fourth indicator

has an outer loading of 0.576. After removing that indicator, the internal consistency and

convergent validity of TTF got significantly better (Table 4), therefore we decided to remove

TTF4 from the outer model.

Table 4 - Indicators before and after removing the items with the lowest outer loading below

0.708

54

Indicators Outer Loadings

Outer Loadings

Cronbach's

Alpha

Composite

Reliability

Average Variance

Extracted (AVE)

TTF1 .88 .896 .772 -> .813 .857 -> .89 .605 -> .731

TTF2 .882 .91

TTF3 .733 .749

TTF4 .576 removed

ENV1 .756 .776 .762 -> .737 .837 -> .83 .508 -> .551

ENV2 .793 .804

ENV3 .656 removed

ENV4 .662 .683

ENV5 .686 .699

USE1 .698 removed .743 -> .728 .837 -> .845 .563 -> .645

USE2 .774 .755

USE3 .753 .82

USE4 .774 .832

Even though the construct Environment has three indicators with outer loadings

below 0.708, if we remove them all, the construct loses internal consistency. We could

consider removing the third indicator, which has the lowest outer loading but internal

consistency and reliability decreases (Hair et al., 2016). Therefore, ENV3 was retained.

Removing USE’s first indicator, which has an outer loading below 0.708, leads to an

increase in the composite reliability and AVE, but a decrease in the Cronbach’s Alpha (Hair et

al., 2016). Therefore, USE1 was retained.

55

In discriminant validity, we are assessing if the constructs in our model are unique and

capture phenomena not present in other constructs (Hair et al., 2016). For that, we are going

to look at the cross loadings. The loading of each indicator needs to be the highest for the

construct they are representing (Hair et al., 2016). In our model all the loadings are higher

than the cross-loadings, which means our constructs are properly measured and distinct from

each other. In table 5, the loadings of each indicator are highlighted in bold for an easier

analysis and comparison.

Table 5 - Indicators’ loadings and cross-loadings.

COMP ENG FC ENV PERF PU VLNT TASK TECH TTF USE

COMP1 0.817 0.43 0.082 0.415 0.553 0.479 -0.09 0.064 0.342 0.451 0.011

COMP2 0.825 0.411 0.283 0.498 0.529 0.461 -0.007 0.149 0.505 0.506 0.273

COMP3 0.881 0.59 0.133 0.535 0.7 0.685 -0.127 0.095 0.446 0.532 0.091

COMP4 0.906 0.533 0.254 0.553 0.689 0.611 -0.028 0.107 0.496 0.622 0.182

ENG1 0.577 0.942 0.105 0.585 0.661 0.524 -0.226 -0.053 0.361 0.421 0.035

ENG2 0.508 0.929 0.203 0.501 0.632 0.447 -0.189 -0.009 0.384 0.367 0.115

FC1 0.202 0.104 0.803 0.276 0.156 0.242 0.197 0.127 0.366 0.298 0.405

FC2 0.136 0.08 0.805 0.252 0.125 0.179 0.241 0.143 0.374 0.297 0.374

FC3 0.14 0.162 0.759 0.266 0.14 0.186 0.09 0.067 0.291 0.252 0.309

FC4 0.207 0.175 0.842 0.243 0.236 0.308 0.18 0.264 0.372 0.335 0.467

ENV1 0.408 0.505 0.152 0.756 0.4 0.358 -0.125 0.03 0.271 0.322 0.064

ENV2 0.584 0.545 0.16 0.793 0.567 0.485 0.006 0.048 0.388 0.496 0.083

ENV3 0.38 0.318 0.291 0.656 0.361 0.3 -0.044 -0.056 0.405 0.329 0.116

56

ENV4 0.376 0.344 0.297 0.662 0.424 0.325 0.055 0.147 0.342 0.304 0.241

ENV5 0.243 0.261 0.341 0.686 0.287 0.233 0.056 0.152 0.348 0.257 0.242

PERF1 0.627 0.539 0.17 0.501 0.863 0.664 -0.091 0.204 0.442 0.531 0.165

PERF2 0.578 0.546 0.056 0.483 0.853 0.633 -0.117 0.15 0.427 0.482 0.102

PERF3 0.676 0.627 0.247 0.523 0.883 0.676 -0.105 0.16 0.527 0.592 0.128

PERF4 0.634 0.647 0.235 0.538 0.847 0.585 -0.104 0.116 0.46 0.523 0.167

PERF5 0.61 0.604 0.178 0.504 0.844 0.647 -0.093 0.203 0.465 0.537 0.135

PU1 0.483 0.34 0.34 0.407 0.54 0.82 -0.055 0.277 0.425 0.554 0.164

PU2 0.635 0.548 0.201 0.467 0.71 0.905 -0.125 0.298 0.462 0.555 0.155

PU3 0.591 0.502 0.209 0.455 0.675 0.907 -0.111 0.258 0.465 0.607 0.193

PU4 0.516 0.391 0.329 0.422 0.588 0.806 -0.065 0.23 0.474 0.575 0.257

PU5 0.626 0.448 0.217 0.408 0.699 0.886 -0.061 0.245 0.471 0.551 0.229

VLNT -0.077 -0.223 0.224 -0.022 -0.119

-0.098

1 0.25 0.022 0.031 0.308

TASK1 0.11 -0.014 0.23 0.08 0.177 0.242 0.203 0.857 0.242 0.237 0.314

TASK2 0.148 0.014 0.145 0.108 0.186 0.298 0.203 0.864 0.195 0.225 0.31

TASK3 0.049 -0.096 0.132 0.022 0.14 0.245 0.248 0.886 0.162 0.204 0.302

TECH1 0.319 0.234 0.388 0.275 0.332 0.397 0.066 0.248 0.758 0.436 0.355

TECH2 0.423 0.338 0.212 0.435 0.38 0.334 -0.04 0.026 0.725 0.463 0.125

TECH3 0.53 0.428 0.305 0.512 0.571 0.519 -0.089 0.159 0.873 0.596 0.246

TECH4 0.358 0.241 0.486 0.297 0.404 0.41 0.143 0.297 0.803 0.546 0.455

57

TTF1 0.486 0.311 0.381 0.395 0.525 0.595 0.083 0.32 0.612 0.896 0.363

TTF2 0.492 0.299 0.361 0.412 0.511 0.554 0.064 0.226 0.552 0.91 0.308

TTF3 0.631 0.498 0.191 0.483 0.571 0.526 -0.086 0.088 0.499 0.749 0.056

USE1 0.236 0.118 0.281 0.157 0.257 0.238 0.206 0.379 0.313 0.281 0.698

USE2 0.158 0.111 0.398 0.156 0.185 0.261 0.135 0.256 0.305 0.233 0.774

USE3 -0.01 0.02 0.33 0.101 0.002 0.078 0.2 0.185 0.191 0.155 0.753

USE4 0.078 -0.01 0.444 0.144 0.038 0.106 0.357 0.241 0.295 0.21 0.774

The Fornell-Lacker criterion uses the square root of the Average Variance Extracted

(AVE) to determine if one construct shares more variance with its indicators than with the

indicators from other constructs (Hair et al., 2016). In our model the square root of AVE for

each construct is higher than the correlation of a construct with all the other constructs.

Therefore, each construct can explain more of its indicators’ variance than the other

constructs can, and so all the constructs are conceptually, and empirically different.

Therefore, discriminant validity is established.

Because there have been studies that consider the cross-loadings and the Fornell-

Lacker criterion week for assessing discriminant validity, Henseler et al. (2015) recommend

evaluating the heterotrait-monotrait ratio (HTMT) of the correlations for assessing

discriminant validity. HTMT is an estimate of the correlations between the constructs and it

can be used as a criterion, where a construct with an HTMT above 0.90 or 0.85, if we choose

to be more conservative, lacks discriminant validity (Hair et al., 2016; Henseler et al., 2015).

In using HTMT as a criterion, there were no constructs with an HTMT value above

0.850 (Table 6). This means all the constructs in our model are empirically different from each

other and the indicators only represent the constructs they are measuring.

Table 6 - Heterotrait-monotrait ratio (HTMT)

COMP ENG ENVIR FC PU PERF TASK TECH TTF USE

58

COMP

ENG 0.656

ENVIR 0.675 0.679

FC 0.254 0.197 0.442

PU 0.720 0.579 0.569 0.339

PERF 0.802 0.781 0.683 0.235 0.813

TASK 0.139 0.062 0.154 0.223 0.344 0.221

TECH 0.614 0.475 0.621 0.540 0.614 0.623 0.287

TTF 0.738 0.517 0.614 0.444 0.762 0.729 0.299 0.799

USE 0.252 0.135 0.273 0.611 0.277 0.206 0.447 0.476 0.382

VLNT 0.078 0.240 0.092 0.244 0.101 0.125 0.274 0.119 0.101 0.347

It can also be used as a statistical test for testing the null hypothesis (𝐻0: HTMT ≥ 1)

against the alternative hypothesis (𝐻1: HTMT < 1), which can be calculated using the

bootstrapping procedure and a confidence interval (Hair et al., 2016; Henseler et al., 2015).

We assumed a 90% confidence interval, and if the confidence interval contains the value 1,

then there is a lack of discriminant validity (Hair et al., 2016; Henseler et al., 2015).

We applied the bootstrapping routine as described in the research design. Neither the

lower (2.5%) or the upper (97.5%) bounds of the confidence interval showed values equal or

above 1, in fact they were all below 0.90. So, discriminant validity is established.

7.2. Structural Model

We have confirmed that the indicators are reliable and valid, and the constructs are

different from each other. Now we are going to assess the relationships between the

constructs in our model and determine the predictive capabilities of our model. While

59

examining the relationships between constructs, we are going to assess the relevance and

significance of the relationships in our model and, therefore, confirm or reject the hypothesis

proposed in the research model. We are going to assess the model’s explanatory power,

which is the in-sample predictive power, through the dependent variables’ coefficient of

determination (𝑅2) values and 𝑓2 effect size (Hair et al., 2021). Finally, for assessing the

predictive capabilities of our model, we are going to use the 𝑃𝐿𝑆𝑝𝑟𝑒𝑑𝑖𝑐𝑡 procedure, which

allows us to evaluate the model’s out-of-sample predictive power (Hair et al., 2021).

The path coefficients reveal the truth about the hypothesized relationships among the

constructs, the closer to + 1 or -1 are the path coefficients, the stronger they are, and the

closer to 0 the weaker they are (Hair et al., 2021). Nevertheless, to truly assess if a relationship

between two constructs is significant, we must know the p values, which are calculated

through the procedure of bootstrapping. The bootstrapping routine is applied as described in

the research design. When assuming a significance level of 5%, the p value must be smaller

than 0.05 to conclude that the relationship is significant at a 5% level.

Figure 5 shows the estimates of the path coefficients for each hypothesized path in

the theoretical model and the 𝑅2 value for each dependent variable. Three out of fifteen

hypotheses were rejected, which are represented by the dotted arrows.

Figure 6 - Structural model results

60

The results show us that Task-technology Fit is significantly and positively impacted by

Task Characteristics and Technology Characteristics, therefore hypothesis H1 and H2 were

supported. Contrary to expectations, Facilitating Conditions does not act as a moderator of

the relationship between Task Characteristics and TTF, thus hypothesis H3 was not supported.

Nevertheless, Facilitating Conditions acts as a moderator of the relationship between

Technology Characteristics and TTF and has a positive significant impact, so hypothesis H4 is

supported.

Task-Technology Fit has a direct significant and positive effect on Individual

Performance and Utilization, therefore hypothesis H5 and H6 were supported. Voluntariness

acts as a moderator of the relationship between TTF and Utilization, having a negative

significant impact, and so hypothesis H7 is supported. Utilization is also significantly and

positively impacted by Facilitating Conditions, thus hypothesis H8 is supported. On the other

hand, Utilization does not have a significant impact on Individual Performance, therefore

hypothesis H9 is not supported.

Engagement has a significant positive impact on Individual Performance, hence

hypothesis H10 is supported. On the contrary Environment does not have a significant impact

on Individual Performance, thus hypothesis H11 is not supported. However, it has a positive

significant impact on Engagement, therefore hypothesis H12 is supported.

Individual Performance is significantly and positively impacted by Compatibility and

Perceived Usefulness, so hypothesis H13 and H14 are supported. Lastly, Perceived Usefulness

has a positive significant effect on Compatibility, thus hypothesis H15 is supported.

Now that we have assessed the relationships’ significance, we need to examine the

relevance of the significant relationships. For that, we are going to look at the total effects

that account for the direct effect and the indirect effect of a variable on another. This is going

to allow us to analyze the two proposed mediation models and discover other unexpected

indirect effects.

7.3. Total Effects

The results showed that most total effects were derived only from direct effects,

however, there were three indirect effects that were relevant at a 5% level.

61

Table 7 shows the total effect, the direct effect, the indirect effect, and their

corresponding p value.

Table 7 - Total effects and direct effects of the significant indirect paths in our model

Total Effect

P value Direct Effect

P value Indirect Effect

P Value

PU -> COMP -> PERF .495 .000 .345 .000 .150 .000

ENVIR -> ENG -> PERF .242 .000 .068 .161 .174 .000

TECH -> TTF -> USE .099 .018 - - .099 .018

This further supports PU as the most relevant predictor of PERF in our model. It has

the most significant direct effect on PERF, but also a significant indirect effect. Therefore, the

total effect of PU on PERF is partially due to its direct effect and partially due to its indirect

effect.

PU’s indirect path is composed by a single mediator, which is COMP. Thus, that

relationship can be read as: PU, which is the independent variable, has a positive significant

effect on COMP, which is the mediator variable. Consequently, the mediator variable COMP

has a significant and positive effect on the dependent variable, which is PERF. This means

hypothesis H17 is supported, and because the direct effect between PU and PERF is positive

and significant, meaning the indirect effect and direct effect are both positive and significant,

there is a complementary mediation (Hair et al., 2021; X. Zhao et al., 2010).

ENV’s indirect path is composed by a single mediator, which is ENG. That relationship

can be read as: ENV, which is the independent variable, has a positive significant effect on

ENG, which is the moderator variable. Consequently, the moderator variable ENG has a

significant and positive effect on the dependent variable, which is PERF. This means

hypothesis H16 is supported, and because the direct effect between ENV and PERF is

insignificant, meaning only the indirect effect is significant, there is an indirect-only mediation

(Hair et al., 2021; X. Zhao et al., 2010).

Even though it was not theorized, Technology Characteristics has a significant positive

indirect effect on Utilization, mediated by one single mediator, which is TTF.

62

7.4. Explanatory Power

Our model’s explanatory power is going to be assessed based on the coefficient of

determination (𝑅2value), which represents the predictors’ combined effects on PERF. This

means PERF is 73.1% explained by all the variables linked to it. When Goodhue and Thompson

(1995) tested their original model, they obtained an adjusted 𝑅2value of 0.16 on the variable

PERF, which was predicted by two variables: TTF and USE, where TTF had a greater impact.

McGill and Klobas (2009), were the first researchers to test the TTF theory in the context of

online education and they obtained an 𝑅2value of 0.448 on the variable perceived impact on

learning, which represented the students’ perception of their PERF. Students' PERF was

directly predicted by TTF and Utilization, but TTF had a stronger impact. In our model, the

𝑅2value is much higher, meaning our model has a great in-sample predictive power,

compared to these studies, but we also must keep in mind that we have more endogenous

constructs leading to PERF, which will naturally increase the 𝑅2 value of PERF. The turning

point is that we have found variables that lead to a greater predictive power. To assess their

importance compared to the classic TTF constructs we are going to need to quantify the

strengths of the relationships between the exogenous and the endogenous constructs.

The 𝑓2 (effect size) is going to quantify the strength of the relationships in our model.

Its value basically represents the decline in the 𝑅2 value when the effect of an explanatory

construct is excluded from the model (Hair et al., 2021).

Table 8 - 𝑓2 of the significant paths in our model (significance at the 5% level).

𝑓2

PU -> COMP .787

ENVIR -> ENG .513

TECH -> TTF .487

PU -> PERF .195

ENG -> PERF .183

63

FC -> USE .138

COMP -> PERF .079

TTF -> USE .035

TASK -> TTF .021

TTF -> PERF .019

USE -> PERF .001

For assessing the effect size, we are using Cohen’s (1988) guidelines, as suggested by

Hair et al. (2021). The values 0.02, 0.15, and 0.36 represent, respectively, a small, medium,

and large effect. When the effect size is less than 0.02, there is no measurable effect.

This means, as can be seen in Table 8, USE, if removed from the model would have no

effect on the 𝑅2value of PERF, and TTF would have a minimal impact, therefore their

contribution to explain PERF ends up being not measurable and very small. PU and ENG have

a medium effect size on the 𝑅2value of PERF, COMP has a small effect size on PERF, and the

strongest effect sizes in our model are the effect of PU on COMP, the effect of ENV on ENG,

and the effect of TECH on TTF, and they all constitute a large effect size.

7.5. Moderation

Facilitating Conditions interacts with two other exogenous variables, separately.

Therefore, there is a two-way interaction.

The interaction term represents the effect of the interaction between TASK and FC on

TTF. It basically shows how the effect of TASK changes when FC decreases or increases by one

standard deviation unit. In SmartPLS, to compute the interaction term, we chose the two-

stage approach with standardized data, since there is evidence that this is the most efficient

way for conducting moderator analysis (Becker et al., 2018; Hair et al., 2021).

As it was discussed before, the effect size helps us understand the true importance of

a relationship between two variables. However, for a moderating effect, the values for the

64

interpretation of the 𝑓2 are different. Effect sizes between 0.005 and 0.010, are considered

small, effect sizes between 0.010 and 0.025, are medium and effect sizes greater than 0.025

are considered large (Hair et al. 2021).

Table 9 - The moderators’ effect size (𝑓2)

𝑓2

TASK*FC -> TTF .007

TECH*FC -> TTF .013

TTF*VOLUNT -> USE .033

Looking at the 𝑓2, by following Kenny’s (2018) guidelines, as suggested by Hair et al.

(2021), FC as a moderator of TECH has a medium effect size (.013) on TTF, and VLNT as a

moderator of TTF has a large effect size (.033) on USE. Even though hypothesis 3 was not

confirmed, if we had accepted a less conservative P value, that is a P value of 0.1, it would

have been accepted. Furthermore, if we look at the effect size, there is still a small effect size

(.007) of TASK moderated by FC on TTF. Therefore, this relationship shouldn’t be totally

discarded, especially in future research.

7.6. Predictive Power

A relevant model is a model capable of producing generalizable findings, so it can be

applied to different samples and reach similar results. To assess that, we are going to use the

𝑃𝐿𝑆𝑝𝑟𝑒𝑑𝑖𝑐𝑡 procedure, from the SmartPLS software. There are three choices we must make

when running 𝑃𝐿𝑆𝑝𝑟𝑒𝑑𝑖𝑐𝑡. First, we must focus on the number of folds. A 10-fold cross-

validation will split our 295 observations into 10 similarly sized groups of data, which means

9 (i.e., k-1) of those groups are going to be combined into a single training sample and the

excluded group is going to work as the holdout sample. This process is going to be repeated

10 times until all groups of data have been excluded one time. Each time, the training sample

is going to be used to predict the holdout or excluded sample. In the end there are going to

be 10 predicted holdout samples and their accuracy is summarized in the prediction statistics.

65

We have 295 observations in our model. If we use a 10-fold cross-validation, our

training sample is going to be approximately 266, which is still within the minimum sample

size requirements, so we can follow the conventional number of folds, that is 10 (Hair et al.,

2021).

Second, we have to decide the number of repetitions, which means we have to decide

how many times the process described above is going to be repeated to produce an average

of predictions from multiple estimations. 10 is the conventional number of repetitions that

we are going to adopt (Hair et al., 2021; Shmueli et al., 2019). After the iterative process is

finished, we are going to look at the prediction statistics which allow us to assess our model’s

predictive power.

First, we are going to look at the 𝑄𝑝𝑟𝑒𝑑𝑖𝑐𝑡2 . All values should be above 0, otherwise PLS-

SEM based predictions cannot even outperform the most naïve benchmark (Hair et al., 2021;

Shmueli et al., 2019). Since they are all above 0, there should be some level of predictive

relevance. This means the PLS path model’s prediction error is smaller than the prediction

error given by the most naïve benchmark.

Next, when the PLS prediction errors are symmetrically distributed we should compare

the Root Mean Square Error (RMSE) values from the PLS path model’s predictions with the

linear regression model (LM) benchmark for each indicator (Hair et al., 2021; Shmueli et al.,

2019). Only utilization’s indicators have skewed PLS prediction errors, and because we are

focusing on Individual Performance’s predictive power, we are going to use the RMSE values

for comparison.

Table 10 - 𝑃𝐿𝑆𝑝𝑟𝑒𝑑𝑖𝑐𝑡 results

PLS RMSE LM RMSE

PERF3 1.227 1.244

PERF4 1.328 1.333

PERF5 1.347 1.395

PERF2 1.441 1.471

66

PERF1 1.329 1.354

Table 10 shows that all PERF’s indicators in the PLS path model have lower RMSE

values compared to the naïve LM benchmark. Therefore, Individual Performance has a high

out-of-sample predictive power (Hair et al., 2021; Shmueli et al., 2019).

8. Qualitative Approach Design: Data Collection Methods

In qualitative research, demonstrating validity is enough to establish reliability

(Venkatesh et al., 2013). Venkatesh et al. (2013), have established three types of validity for

qualitative research, which are: design validity, referring to the goodness of the design and

execution of the interviews to produce credible and transferable findings; analytical validity,

which is related to the goodness of the collection and analysis of data to produce dependable,

consistent, and plausible findings; and inferential validity, which is the quality of the

interpretation, which means our findings can be confirmed and corroborated by previous

studies.

8.1. Interviews

To collect primary data, we conducted telephone semi-structured interviews. The

interviews were conducted during the end of October and the beginning of November 2021.

We opted to collect data this way, due to the constraints created by the COVID-19 pandemic.

This seemed a viable (Novick, 2008) and safe way that would be easy and familiar for the

participants. The interviewees were placed on speaker and the conversation was recorded

with their consent.

To ensure there would be no ethical issues in the process of collecting data, the

purpose and title of the study were disclosed to the participants, as well as the identification

of the researchers and institutions they belong to. It was explained to the participants why

they were selected, and their confidentiality was guaranteed. Also, we made our contacts

available, in case there were questions from the participants (Creswell, 2009).

67

8.2. Sample

This study was conducted in Portugal and targeted students attending courses at

different Lisbon universities. Since we are conducting a sequential mixed methods study, we

adopted a sequential sampling strategy, which is based on the methods and results applied

in our first strand of research. Being so, we conducted our qualitative study using a parallel

sample, which means our qualitative sample is not the same as our quantitative sample but

it was drawn from the same underlying population (Collins et al., 2007; Venkatesh et al.,

2013).

Therefore, to further test the predictive power of our model, we interviewed 8 college

students, attending universities in Lisbon, that had not participated in the survey from our

quantitative study, including two students from two different universities which were not

present in our list of surveyed universities.

8.3. Sample Characteristics

The universities included in our qualitative study are the Faculty of Medicine (FMUL)

from the University of Lisbon, the Faculty of Human Kinetics (FMH) from the University of

Lisbon, the NOVA School of Social Sciences and Humanities (FCSH) from the NOVA University

of Lisbon, the Superior Technical Institute (IST) from the University of Lisbon, and the

Lusófona University of Humanities and Technology (ULHT) from Lusófona University of Lisbon.

Table 11 – Interviewed students characteristics

Identification University Course Gender Age

Student 1 FMUL Medicine Female 23

Student 2 FMUL Medicine Female 21

Student 3 FMUL Medicine Female 22

Student 4 FMH Sports Sciences Male 24

Student 5 FMH Sports Sciences Male 21

Student 6 FCSH Geography Male 22

Student 7 IST Engineering Male 21

Student 8 ULHT Cinema Male 21

Among the 8 interviewed students, there were 3 females and 5 males. Most

participants were solely students for their occupation and were living with their parents,

68

which means they could focus on learning and had very few other occupations to balance

with college. Only Students 4 and 8 had a part-time job during the pandemic, however being

a student was their primary occupation. Most students were 21 years old but the ages ranged

between 21 and 24 years old.

8.4. Data Collection and Measurements

To ensure design validity, we based our interview questions on our questionnaire

questions, which were already validated and deemed of quality, however we converted them

into open-ended questions. Also, we included extra open-ended questions in order to further

explain the findings from our quantitative study and to answer our research questions.

To support our semi-directed telephone interviews, we developed a script to help

guide the conversation, and facilitate a coherent flow of information (Castillo-Montoya,

2016). We first tested the script with two students, and after perfecting and organizing our

interview flow, we proceeded to conduct the interviews with the 8 students.

We interviewed the individuals separately and each interview was audio recorded. We

started the interviews with introductory questions to ease the participants into the

conversation and make them more comfortable with describing their experience (Castillo-

Montoya, 2016). With those questions, we also intended to distinguish the interviewees.

After the introductory questions we transitioned to our key questions where we collected the

most valuable information to answer the research questions (Castillo-Montoya, 2016). We

finished the interviews with two closing questions, for the participants to be able to reflect

and add information and to take the opportunity of getting to know issues not addressed

before (Castillo-Montoya, 2016).

During the interview we made sure the interviewees understood our questions, and

in return that we understood their answers. For transcribing the interview, we used

Transkriptor (Transkriptor, 2021), which is an online tool for transcribing audio recordings.

After transcribing the interview, we proofread the transcribed audio recording to make sure

there were no mismatches between the audio and the text. This was done to ensure credible

and transferable findings (Venkatesh et al., 2016).

69

9. Qualitative Approach Design: Data Analysis

The qualitative study aims to further support and explain the findings from the

qualitative study and gain knowledge from the interviewed students’ detailed views. This

study aims to discover how the functionalities of collaborative platforms can limit or benefit

students and allow them to complete their tasks and how they can impact their individual

performance. Furthermore, we want to support the findings that go against the status quo in

the literature, regarding the relationship between USE and PERF.

After the transcription of our interviews, we organized and prepared our data for

analysis. For that, we read through all the transcriptions, in order to obtain a general sense of

the information and to start coding our data. For coding our data, we used MAXQDA 2020,

which is a software for text analysis. The first code was utilization, which was subdivided into

platforms used, most used, utilization frequency, voluntariness, utilization’s impact on

individual performance, and future use. The other codes were directly related to the

constructs in our structural model, which were: facilitating conditions, technology

characteristics, task characteristics, TTF, perceived usefulness, compatibility, engagement,

environment, and performance. We also included two codes related to the positive attributes

and negative attributes of collaborative technologies to help complete the information on the

other constructs.

10. Qualitative Results

This chapter presents the detailed results of the content analysis of our transcribed

interviews using the software MAXQDA 2020 version 20.4.1 (VERBI Software, 2019). The

presentation of our qualitative results is going to be sectioned by the codes’ names. For each

section, first, we are going to present and analyze the perspectives of each student on each

subject. Second, we are going to conduct a descriptive analysis of the questions in the

questionnaire that inspired the questions in our interview.

For each section, each item measuring a construct in the questionnaire, is represented

in a table with the indicator they represent and their mode. The responses were divided in

three categories, which are: Negative, where the responses totally disagree, disagree, and

somewhat disagree, were enclosed to represent students with a negative perspective;

Neutral, where students’ that responded neither agree nor disagree were enclose; and

70

Positive, where the responses totally agree, agree, and somewhat agree, were enclosed to

represent students with a positive outlook.

10.1 Utilization

10.1.1. Platforms used, most used, utilization frequency, and Voluntariness

Student 1 used Zoom for video conferencing and Moodle for online learning, accessing

pdfs, and classes’ slides. Student 2 used Zoom only, as well as Students 3 and 5 and they all

used it every day, several times a day. Student 4 used Skype, Zoom, Blue Button and Google

Drive, but mainly Zoom, which was used every day several hours a day. Student 6 used Zoom,

Google Meets, and Microsoft Teams but mainly Zoom, which he used every day for online

classes. Student 7 used Zoom, Discord, Microsoft Teams and Slack, but he mainly used Zoom

for daily online classes. Student 8 used Zoom, Microsoft Teams, Google Docs, and Google

Drive, but mainly Zoom, for daily online classes.

For all students, the most used platform was Zoom. All students had a high frequency

of utilization, meaning they utilized collaborative platforms several hours a day during the

week, for academic purposes. During the interviews, all students refer to utilization of Zoom

for online classes, group projects, and evaluation elements. Furthermore, all students felt that

Collaborative Platforms’ utilization was mandatory, meaning they felt their teachers required

them to use collaborative platforms for online learning. However, even though they felt that

that was their only option, they also understood it was a necessity during the pandemic, and

that it was the most viable way for communicating with others, especially for working on

group projects, not just online classes.

Table 12 – Questionnaire responses to utilization and voluntariness’s items

Item Indicator Mode Negative Neutral Positive

I have already used

collaborative platforms for

communicating matters

related to projects.

USE1 7 8% 10% 82%

71

I have already written and

read messages or comments

on collaborative platforms.

USE2 7 4% 16% 90%

I have already used

collaborative platforms to

attend classes.

USE3 7 1% 2% 96%

Currently, I frequently use

collaborative platforms

USE4 7 2% 3% 95%

My teachers require me to

use collaborative platforms.

VLNT 7 7% 18% 75%

When we look at the data from our surveys, we can understand that the 295

observations had in great part a high level of Utilization. The data tells us that 99% of the

surveyed students used Zoom for distance learning during the COVID-19 pandemic but

simultaneously 30% used Google Tools and 12% used Microsoft Teams. It is interesting to

note that the 8 interviewed students also had a high level of utilization and the platform they

all used the most was Zoom. In terms of Voluntariness, most of the surveyed students felt

that the use of collaborative platforms was something required by their teachers. Which was

the same case for the interviewed students. This highlights that both samples reflect each

other, even though they are different.

10.1.2. Utilization’s impact on individual performance

Student 1 does not think an increase in use means an increase in individual

performance. In the beginning the novelty factor made the utilization of collaborative

platforms fun. However, the interviewee considers that with time, use ends up having a

negative impact on her performance since a high level of use endangers her ability to focus

and her quality of life. Similarly, Students 2 and 5 are certain that increased use does not

relate to their individual performance as students. They all felt use was mandatory.

Student 3 thinks increased utilization can lead to an increased impact on Individual

Performance but that depends on the technology. For example, the increased utilization of a

computer or a mobile phone can lead to a greater performance, whereas “(...) when it comes

to Zooms and things like that, no, I don’t think so.” Furthermore, Student 3 feels that she

72

needs to experience the use of a technology she has never used before to assess if it can have

a positive impact on her performance, “For a mobile phone and a computer I am already

expecting a positive impact” since those are technologies she has experienced before.

Similarly, to other students, Student 3 feels the use of Zoom was mandatory because the

classes were mandatory, so there was really no choice to be made.

Student 4 does not think an increase in use is related to an increase in his performance.

However, "what counts is (...) going to more classes because the more classes I go to the

better (...) my performance will be. He also thinks there are other factors that play a bigger

role in his performance, such as: "(…) being focused during classes and studying at home just

like in normal classes, it is not because of the use of the platform per say." Transmitting the

idea that the platforms he used worked more like a vehicle for students to perform their

normal functions as students, it was a means to an end. Student 4 thinks use was mandatory

because of the conditions at the time, because of the conditions at the time of the COVID-19

pandemic.

Student 6 does not feel that the use is directly related to performance, he thinks an

increase in use makes the companies who build the platforms to "(…) feel the need of evolving

them to satisfy the needs of its users." Furthermore, he thinks collaborative platforms were

an essential help for all the students around the world to perform their functions as students.

Also, when he used collaborative platforms for the first time, he was already expecting a

positive outcome, however he needed to test it once to assess if that was true. Student 7 felt

the use was mandatory because in his university Zoom was the only collaborative tool allowed

for online classes, "(…) so in that field I had no choice."

Student 7 was already expecting collaborative platforms to perform well, when he

started to use them, and for him they always did. As he says, "I don't think I have ever

questioned their functioning”. Also, he feels that an increase in use does not lead to a better

performance, he feels his "(…) performance was always the same" and that there was no

improvement. Furthermore, he felt that the use was not voluntary, but he also did not mind

it because "it was the only alternative that we had at the time, and I think it was the best

alternative we had at the time."

73

Student 8 feels that an increase in use creates habit and makes working with the

platform more intuitive and quicker. But that can also be a disadvantage since people become

too used to working on Zoom and avoid meeting in person. Furthermore, the student feels

like there is a trial period when using a new technology to assess if they are actually useful or

if he should continue to do things the old way. Student 8 did not feel that using the platform

was against his will because he felt it was established as a norm that Zoom was the best way

to attend classes, so he obliged and did not oppose it. In his own words "in a way it was

mandatory, but a necessary mandatory."

10.1.3. Future use

All students except for two students would like to be in a blended learning regime for

the new academic year. Student 4 would like to follow the traditional face-to-face learning

method but with the possibility of class recordings, so he could rewatch them at home. He is

now taking a master’s degree and that method of learning is already being employed, as his

professors felt that since the recording of classes worked so well during the pandemic, it

would be of their interest to continue class recordings, so students could continue to achieve

better learning performance. Student 5 is the only interviewed student preferring the face-

to-face learning method. As we are going to understand further ahead, student 4 is the

student with the lowest levels of compatibility, PU, and TTF.

10.2. Facilitating Conditions

Student 1 has a high level of Facilitating Conditions since she felt she had all the

necessary resources to utilize Collaborative Platforms, such as a computer and good internet

connection and the necessary knowledge, such she characterized Zoom as user-friendly.

Students 2 and 5 also have a high level of Facilitating Conditions because they have

the necessary resources, knowledge and help when needed.

Student 3 thinks she has the necessary resources to use Zoom, such as a computer

with a camera. When it comes to the necessary knowledge, she doesn't feel the most

confident because she is “(...) not very good with working with computers and things like

that”, even though she was able to use them. Student 3 could easily get help from others

when experiencing difficulties, she would usually ask “(...) on the chat like privately or call

someone.”

74

Similarly, Students 4 and 8 have the necessary resources such as a computer and a

good internet connection. They had a bit of trouble when first using Zoom “(...) but after the

adaptation it was fast and easy” and “(...) it was a bit difficult to understand at first, but I

quickly understood and after, I knew how to use the program quite well.” They also had help

from colleagues and teachers when there was a problem in using Zoom.

Student 6 considers that the use of Zoom is easy, since “it was just accessing the link

and we could soon watch the class.” Furthermore, if there was any difficulty, his colleagues

would quickly help “and that always facilitated it.” Student 6 also feels you do not need much

to attend classes via Zoom, thus “with a good connection and (...) a minimally adequate

computer, I think you can make the most of classes.”

Student 7, like all the other students before, has a high level of Facilitating Conditions.

He has all the necessary resources and knowledge and could easily get help from his

colleagues and professors when needed.

Table 13 - Questionnaire responses to facilitating conditions’ items

Item Indicator Mode Negative Neutral Positive

I have the necessary resources to

use collaborative platforms.

FC1 7 2% 6% 92%

I have the necessary knowledge to

use collaborative platforms.

FC2 7 2% 5% 93%

I can get help from others when I

have difficulties using

collaborative platforms.

FC3 7 4% 10% 86%

collaborative platforms are

compatible with other

technologies I use.

FC4 7 2% 8% 91%

When we look at the data from our surveys, most of the 295 students had a high level

of Facilitating Conditions. The data tells us that more than 90% of the surveyed students had

the necessary resources and knowledge to use Zoom and all the other platforms they used

for distance learning during the COVID-19 pandemic. 86% of the surveyed student were

positive about the help they could get when experiencing difficulties with collaborative

75

platforms. It is interesting to note that the 8 interviewed students also had a high level of FC.

This highlights that both samples represent each other, even though they are different, and

further supports and explains our quantitative data.

10.3. Technology Characteristics and Task Characteristics

Student 1 considers Zoom’s functionalities useful and user-friendly. When it comes to

functionalities of collaborative platforms, for Student 1, the most useful are: the chat,

because she feels it gives more courage to students for asking questions to the professor,

when in a face-to-face environment they would not, and the polling feature. Also, meetings’

recording, which was very helpful for having classes on-demand, as students could rewatch

lectures whenever they wanted and at their own pace. On the other hand, the final exams

were not done through the platforms because of data protection and privacy matters. This

was an unexpected problem for Student 1 because the exams of 2 semesters were postponed

to the end of the year. In terms of academic tasks done through collaborative platforms,

Student 1 had to do group projects and attend mainly theoretical classes. There were only

“(…) one or two visits to patients filmed by the professors.”

Student 2 also considers Zoom’s functionality useful, but the most useful are the chat

and meetings’ recording, especially for rewatching the recorded lectures. As a user, Student

2 did not experience any troubles with the platform, and she would rank high the overall

quality of collaborative platforms. Student 2 had to perform academic tasks via collaborative

platforms such as, group projects, meetings for those projects, discussing medical histories,

and attending meetings for extracurricular projects related to the faculty. Therefore, real-

time communication was very important because it avoided commutes.

Student 3 gives a good rating to the overall quality of collaborative platforms. She has

used functionalities such as instant messaging, meeting recording, screen sharing, breakout

rooms, and the white board on Zoom but she feels the most useful functionalities are screen

sharing, so students can see PowerPoint presentations and screen recording for recording

lectures to be rewatched later. However, there were some unexpected problems in the

beginning. For instance, there were classes with about 370 students, so not all students could

attend the class via Zoom, so they would transmit it on YouTube, at the same time. However,

soon that problem was solved when the faculty figured they could add more participants to

76

the Zoom Pro accounts using the 'Large meeting' add-on. Student 2 had to frequently

collaborate with colleagues since “(...) normally the projects in my course are all in groups,

there are not many individual things.” Through collaborative platforms, she would attend

classes online, mostly theoretical, and about once a week, there would be a practical class

and project meetings. Thus, Student 2's need for real time communication is clear.

Student 4 considers Zoom’s functionalities very useful. Screen sharing, meeting

recording, and the breakout rooms were the best features, in his opinion. The only problem

he experienced was related to the internet connection and therefore, he valued the overall

quality of collaborative platforms as really good. Student 4 used collaborative platforms for

evaluation tasks, such as tests and presentations, group projects, oral presentations, and

written papers. Thus, he felt the need to communicate with others real-time because “(...) it

makes much more sense for all of us to be present, even though not physically, (…) discussing

the projects.” Online classes followed the traditional model implemented before the COVID-

19 pandemic, with the teacher speaking to an audience with the support of a PowerPoint

presentation.

Student 5 thinks Zoom has a lot of useful functionalities. He felt the large capacity for

meeting participants was helpful, because it allowed for the entire class to participate in the

meeting, the breakout rooms, the raise hand button, and meetings’ recording. Student 5 did

not experience unexpected problems with Zoom, except for sporadic problems with the

internet, as he has a good internet connection. Overall, Student 5 thinks Zoom is an overall

good quality collaborative platform, and he used it to attend online classes, do group projects,

and present oral presentations. He mostly had theoretical classes but also some practical

classes.

Student 6 considers the functionalities of collaborative platforms useful but the most

useful are screen sharing and meeting recording, so he can rewatch the online classes.

Furthermore, he did not experience unexpected problems with Zoom. Student 6 utilized

collaborative platforms for tasks such as online classes, group projects, and oral

presentations.

Student 7 considers Zoom’s screen sharing and the TeamViewer’s remote desktop

feature, the most useful functionalities. He did not experience unexpected problems during

77

the utilization of those collaborative platforms, he regards them as useful for real-time

remote communication and “they all function how they are supposed to function.” Student 7

used them for tasks such as online classes and group projects.

Lastly, Student 8 regards Zoom’s screen sharing functionality essential because in the

cinema course’s editing department, which he is part of, students and teachers need to share

what they are doing on the computer. Student 8 experienced unexpected problems because

some classes were split between 40 minutes, the maximum meeting time for the free plan,

so all students would have to reenter the meeting every time it ended. Overall, Student 8

considers Zoom’s functionalities useful, and classifies the overall quality of all collaborative

platforms he used really great, “they all function really well, and they all helped a lot.” Student

8 mentions he used collaborative platforms for online classes, but they were especially

important for group projects, since his course has mostly group projects. For that, besides

using Zoom, he also used Google Docs, Google Drive and WhatsApp.

Table 14 - Questionnaire responses to technology characteristics’ items

Item Indicator Mode Negative Neutral Positive

Collaborative platforms

provide real-time

communication.

TECH1 7 4% 5% 91%

Collaborative platforms had no

frequent unexpected

problems.

TECH2 5 36% 17% 47%

I would give an excellent rating

to the overall quality of

collaborative platforms.

TECH3 5 15% 19% 66%

Overall, collaborative

platforms have useful

functionalities such as instant

messaging, video recording,

screen sharing, etc.

TECH4 7 3% 4% 94%

When analyzing the survey responses, we can determine that most of the 295

surveyed students considered collaborative platforms to provide real-time communication

78

and, overall, a useful set of functionalities. When it comes to the overall quality of

collaborative platforms only 66% of the students gave a positive rating, 19% felt neutral, and

15% gave a negative rating to the overall quality of collaborative platforms. With a less

unanimous response, 47% of the surveyed students were positive about not experiencing

frequent unexpected problems with collaborative platforms. On the other hand, 36% of the

surveyed students felt negative about that, and 17% felt neutral.

Table 15 - Questionnaire responses to task characteristics’ items

Item Indicato

r

Mode Negative Neutral Positive

Frequently, I had to collaborate

with others in my academic tasks

TASK1 6 15% 11% 74%

To perform my academic tasks, I

need to communicate with

others anytime and anywhere.

TASK2 5 21% 15% 64%

My academic tasks required

frequent coordination with the

efforts of others.

TASK3 5 19% 18% 64%

74% of the survey respondents were positive about having to collaborate with others

in their academic tasks, during online learning. 64% of the 295 surveyed students were

positive about their need to communicate with others anytime and anywhere and about

requiring frequent coordination with the efforts of others.

10.4. Task-Technology Fit

Student 1 thinks collaborative platforms fitness “(...) depends on the content of

classes”, therefore she considers them very adequate for theoretical classes but the extreme

opposite for practical classes. As Student 1 states: “if they were practical classes via Zoom (…)

I think it’s useless! I think it’s like a 1. Now, for theoretical classes, I think it’s like a 7.” During

the time Student 1 was having online classes, most of the classes were theoretical classes,

and for those classes Student 1 values Zoom’s real-time communication and recorded classes

because before the COVID-19 pandemic, she would not have attended them, and in this way

she did. For group projects, Student 1 regards Zoom as very adequate since it helped

79

conciliate students’ availability and schedules. For group projects, “(…) it is a new tool that

came to stay (…)” because it can help disperse people in terms of location to meet, improving

quality of life, as it avoids commutes. But on the downside, it hinders medical students’

education because the experimental and practical classes are hard to experience through

Zoom, since there is no real contact with patients or the hospital. The exams also had to be

postponed to the end of the year because the university felt they were not fit for online

collaborative platforms.

Student 2 considers the functionalities of collaborative platforms suitable and enough

because her course tasks were adapted to fit them. Nevertheless, she feels her learning

requirements were only partially met as “(...) it obviously does not have the practical

component (…)”, so “what was once practical ceased to exist”. Because of that, everything

became very theoretical, even with the support of collaborative platforms “(...) what ended

up prejudicing me.” However, the theoretical part of her course fit very well with the

functionalities of collaborative platforms.

Student 3 feels the same way about the collaborative platforms used for online

learning. She thinks they only have adequate functionalities for group projects, meetings, and

theoretical classes. Furthermore, she feels students actually benefited more from theoretical

classes via Zoom. Due to the classes' recordings “(…) we managed to study the subjects better

because many times the teachers don’t give us that much support (…) and there, we have the

recorded lectures (…)”. However, she feels that practical classes via Zoom were not good

enough, as “you can’t train doctors without in-person contact with patients”, since “(...) we

need to be with the patients, talk with the patients, and perform maneuvers, etc. (…)” and

“(...) via Zoom, it was more like a conversation and creating hypothetical situations. There was

no real contact.” Therefore, she felt her learning requirements were not being met since the

practical component of her course was put aside; “(…) it became very incomplete”. That is

why, in the second mandatory confinement, medical students were allowed to have some

practical classes in the hospital.

Student 4 thinks collaborative platforms “(...) can supply one part, which is the most

theoretical part, but taking into account that my course is a very practical course, they could

not satisfy that part.” He also regards that this was not due to the lack of functionalities in the

80

platforms, as “(...) it is impossible for them to satisfy that more theoretical part because that

part really requires us to be in-person (…)”.

Student 5 thinks collaborative platforms are not enough for all aspects of his learning

requirements, as his course has a strong practical component, and all practical classes were

canceled or substituted by less effective alternatives. Student 5 also thinks they were not fit

for evaluations, such as tests and oral presentations. However, he feels they were fit for

theoretical classes especially because of the recorded sessions, which are “very useful for

studying or revising (…)” Student 5 also considered them fit for doing group projects and

attending meetings.

Student 6 is in a more theoretical course so his perception on TTF is very positive. He

feels collaborative platforms had an adequate set of functionalities for his academic tasks, “I

don’t feel that I was harmed. I think that in-person there existed some hypothesis, for

example, in my course we used a lot of maps, but they just started to be presented in a digital

format (…) That I have felt an actual difference, I think that is the only thing.”

Student 7 considers collaborative platforms adequate and with enough functionalities

for the completion of his academic tasks. His course has a more theoretical component, as his

practical classes consist of resolving exercises and problems. Therefore, he feels “there is no

impediment at that level”, meaning he felt his learning requirements were met by the

platforms.

Student 8’s academic tasks are strongly supported by the use of computers. Therefore,

he felt collaborative platforms were a good fit for completing his academic tasks, as they allow

for screen sharing. However, he affirms that it is preferable to be in-person as it is easier for

the professor to understand students’ doubts.

Table 16 - Questionnaire responses to TTF’s items

Item Indicator Mode Negative Neutral Positive

Collaborative platforms provide a

suitable set of functions for my

academic tasks.

TTF1 6 12% 11% 77%

81

Collaborative platforms have

enough functionalities to help me

perform my academic tasks.

TTF2 6 14% 15% 72%

I believe that using collaborative

platforms for online learning can

meet all aspects of my learning

requirements.

TTF3 4 41% 19% 40%

77% of the 295 surveyed are positive about collaborative platforms providing a

suitable set of functions for their academic tasks, and when asked if they think those

functionalities are enough the number drops to 72%. When confronted about their opinion

on the statement “I believe that using collaborative platforms for online learning can meet all

aspects of my learning requirements”, the responses almost behave like a reflection on a

mirror, meaning 41% of the students have a negative perception about this statement and

40% a positive perception. For that statement, most of the surveyed students had a neutral

stance, with 4 being the most frequent response.

10.5. Compatibility

Student 1 does not regard Zoom as compatible with all aspects of her course since her

medical degree has a strong practical component, and Zoom is in nature a video conferencing

platform. Student 1 suggested AMBOSS, which she deemed more compatible, as it was

designed as educational platforms for medical students. Student 1 also does not feel it is

compatible with the way she likes to work because she needs real contact with patients and

that is not possible in an online environment. On the bright side, for meetings and

collaboration, it is compatible with the way she likes to work.

Student 2 feels that the collaborative platforms she used during the pandemic for

online education had nothing specifically designed for her medical degree and they were not

compatible with all aspects of her course. However, they were compatible with the way she

likes to work as the recorded classes enabled Student 2 to better organize her time.

Student 3 prefers to have classes and oral presentations face-to-face. With online

learning, the student feels demotivated, as she does not like to work alone in her room apart

from her colleagues. She also feels mandatory in-person classes demand students to be more

present and involved, whereas in online classes students could turn off their cameras and

82

microphone and just sleep in or go on about their days. Student 3 regards them as compatible

with the situation during the mandatory confinement, as students could not leave their

homes and the hospitals were chaotic. She also feels there was no other option, and it was

the best option at the time. However, it made her education incomplete. Therefore, she feels

her medical course is not entirely compatible with collaborative platforms as her “(...) course

is a practical course, and we have to have real contact with patients and with people (…)”. On

the other hand, she feels they can be compatible with theoretical classes, as they were never

mandatory and so, students could watch at home the recorded sessions and study for the

exams. She also prefers to have meetings via Zoom because it avoids long travels, and it is

more convenient.

Student 4 does not think collaborative platforms are compatible with all aspects of his

course. They are also not compatible with the way he likes to work, as he prefers “(…) to work

in a face-to-face way with people (…)”, but he recognizes it as a good alternative and probably

the best. Finally, “(...) considering the more theoretical aspect of my course, it (Zoom)

managed to satisfy it perfectly.”

Student 5 thinks that “given the circumstances they were quite useful, but it is not the

best option. If it can be in-person, I think it is better to be in person.” He considered them

compatible with group projects and theoretical classes but not for practical classes, exams,

and oral presentations. Overall, he thinks they were compatible with his needs and the way

he likes to work.

Student 6 thinks collaborative platforms are compatible with all aspects of his course,

and they are also compatible with the way he likes to work, as they allow him to multitask

and take better notes during classes. He prefers to do oral presentations virtually as he feels

less nervous, in comparison to an in-person scenario.

Student 7 recognizes that considering the pandemic situation, collaborative platforms

“(…) are the best way we have to learn”. He thinks they are compatible with his course and

are able, to some extent, fulfill his needs as a student. However, they are not compatible with

the way he likes to work and learn as he prefers face-to-face teaching since “(…) it helps us

gain other notions about what we are learning and it helps reinforce what we are learning

83

and therefore, I would prefer not having to use them (collaborative platforms), but we have

no other option and so, it has to be''.

Student 8 considers collaborative platforms compatible with the way he likes to work,

as he enjoys the workflow created by Google Drive and Google Docs. Furthermore, he thinks

collaborative platforms are compatible with his course, his needs, and his current situation.

Even though he thinks students will be more engaged and their doubts and questions will be

better answered in a F2F class.

Table 17 - Questionnaire responses to compatibility’s items

Item Indicator Mode Negative Neutral Positive

Collaborative platforms are

compatible with all aspects of my

course.

COMP1 1 53% 14% 33%

Using collaborative platforms is

compatible with my current situation.

COMP2 5 19% 16% 65%

I think using a collaborative platform

fits well with the way I like to work

COMP3 1 48% 16% 36%

The functionalities of collaborative

platforms serve my needs perfectly.

COMP4 4 35% 20% 44%

We can assess that students' responses to the statements measuring the construct

Compatibility are less positive. Most of the surveyed students (65%) were positive about the

use of collaborative platforms being compatible with their current situation. The other 3

statements measuring Compatibility had fewer positive results. Only 44% of the surveyed

students were positive about collaborative platforms’ functionalities serving their needs and

most of the students answered they neither agree nor disagree with that statement. Only

36% of the surveyed students were positive about collaborative platforms fitting with the way

they like to work, while 48% were negative about it, strongly disagreeing being the most

frequent answer. Lastly, the majority (i.e., 53%) of the surveyed students did not feel

collaborative platforms were compatible with all aspects of their course, with 1 being the

most attributed classification.

84

10.6. Perceived Usefulness

Student 1 felt that in the beginning of her experience with Zoom there was an increase

in her productivity because of the recorded classes. In her eyes Zoom has useful

functionalities that can be useful for learning if they are used for theoretical classes, as she

feels they are useless for practical classes. Also, she feels they are useful as they make

organizing her time easier, allow her to communicate with people who are distant, and

remove the necessity to commute, and thus are useful for group projects and meetings.

Furthermore, they lessen the barriers between students and teachers and promote

communication and participation, especially for shy students. However, it is dependent on a

good internet connection, which can lead to technical difficulties. Plus, it hinders the

relationship between patients and doctors, and promotes distance between people.

Student 2 thinks collaborative platforms are useful tools for learning, but she thinks

they did not increase or decrease her performance, it stayed the same. In her opinion they

are useful because having online classes saves time and they can contribute to a better

conciliation of her schedule. Therefore, “(...) it ends up being more efficient in terms of time,

it does not make us commute and it allows us to have the recordings, which is great for a

student, because if we get dispersed during class, it is possible to rewatch the class (…)”. It is

not as useful as students can’t put into practice what they have learned in class.

Student 3 does not think collaborative platforms contributed to her productivity,

because when she attended classes via Zoom, she would easily disconnect, she would not

participate, and she would have the camera turned off. This would be the norm for classes

with a great number of students. However, when asked if that was more the student’s and

the professor’s fault and not so much the platform's fault, she explained that there were some

classes with fewer students, where they were required to have their cameras turned on, the

professor would interact and question the students, and would promote participation. In

those cases, she would feel the need to be always engaged. Nevertheless, and even though

the teacher tried to make the best of the situation, Student 3 perceives in-person classes a lot

more productive. Student 3 stresses that “obviously the face-to-face regime is always better

but in a pandemic context it is clearly preferable and a lot more productive to have

collaborative platforms than to have nothing (…)”. Also, “(...) in the beginning it seemed like

something innovative (…)”, so it would be more engaging for her to use these platforms.

85

However, her productivity, in the long run, was worse. For Student 3 the most useful thing

about Zoom was screen recordings. However, she was already familiar with this method, as

there were recorded classes, before the COVID-19 pandemic. Also, she had something called

“desgravadas”, which are the transcriptions of audio recorded in-person classes and they

were readily available to students, before the COVID-19 pandemic. Furthermore, she

considers Zoom useful for the pandemic context so people can communicate face-to-face,

and for people who live far away. For group projects, Student 3 experienced some

miscommunication problems, she also felt it was more effective to communicate in-person,

and, in her opinion, when in person people are more productive and don’t drag tasks for so

long. Google Drive, Google Docs, and Google Slides, for group projects, were not a novelty for

the student, and she considers them very useful for group projects.

Student 4 thinks that, considering the context of the mandatory confinement during

the COVID-19 pandemic, collaborative platforms enhanced his productivity. He also feels that

due to that situation, he managed to discover several functionalities that can enhance his

efficiency when doing group projects. Furthermore, he perceives them as useful learning

tools, even in a non-pandemic context, but more for working on projects than for classes.

They are useful as they allow “(...) people to be in the comfort of their homes and still be

learning, lose less time in commutes, communicate with others in real-time (…)” when distant.

However, Student 4 feels they are useless for practical classes, and in case there is a failure in

the system, they become useless for communication.

Student 5 feels collaborative platforms do not enhance his productivity, as in online

learning he frequently got distracted. Therefore, he prefers face-to-face learning as he is more

focused and there are less distractions. Although he thinks they don’t contribute to his

productivity, they were useful during the pandemic context since they allowed students to

have classes at “(...) 70% and if it was not for the platforms, it would be very complicated.”

Plus, in his opinion, they are advantageous as they allow recordings, which are great for

studying and they are also useful for group projects’ meetings.

Student 6 perceives collaborative platforms as useful learning tools, as they support

“(...) better teaching, in a faster way, without so many interruptions”. They were also useful

to prepare Student 6 for the companies’ digital work environment. The only downside, which

86

is not directly linked to collaborative platforms, is a bad internet connection, which can impact

the usefulness of these online platforms.

Student 7 thinks that sometimes it can be complicated to be productive, as he

experienced misunderstandings caused by communicating through collaborative platforms.

Since in-person “(...) we can easily assess what the other person… is trying to say through

body language (…)”. For Student 7 real contact with other people is very important. He also

perceives them as useful tools for learning, since when there is not the possibility of being

there in person, “(...) we always have that means, which helps us obtain supposedly the same

teaching quality we would have in person. That is, excluding the communication part”, which

for him, comparatively is not the same thing. Even though Student 7 thinks communicating

through Zoom can lead to misunderstandings, it was still advantageous for real-time distance

communication. Furthermore, video conferencing humanized the interactions, and the

breakout rooms were great for separating the class in groups. TeamViewer's remote shared

desktop eased processes in group projects. In his opinion, the only downfall is the tools’

dependence on the internet connection, which could lead to lags.

Student 8 perceives that productivity is more connected to someone's personality: “I

think a person has to be naturally productive or get the motivation elsewhere. I don’t think

it’s so much because of having Google Docs or Zoom that a person becomes more productive.

But it helps a little, yes it helps a little.” Furthermore, Student 8 considers using collaborative

platforms “(...) very advantageous (…) in my course I think it’s essential (…)”. As his course is

very computer-based, these tools were all very useful, as students would share files and notes

through Google Drive and would have meetings through Zoom. During classes screen share

and the chat for sharing links among students and teachers, were also perceived by the

students as advantageous and useful.

Table 18 - Questionnaire responses to perceived usefulness’s items

Item Indicator Mode Positive Neutral Negative

Using Collaborative platforms

enables me to accomplish my

academic tasks quickly

PU1 4 18% 27% 54%

87

Using collaborative platforms

enhances my productivity

PU2 3 40% 21% 39%

Using collaborative platforms

makes accomplishing my

academic tasks easier.

PU3 4 27% 29% 43%

I believe collaborative platforms

are useful learning tools

PU4 5 9% 19% 72%

Overall, I find it advantageous

to use collaborative platforms

for my academic tasks.

PU5 5 18% 22% 60%

Most of the surveyed students are positive about collaborative platforms being useful

learning tools (72%) and in being advantageous to use for academic tasks (60%). When

confronted about their opinion on the statement “Using collaborative platforms enhances my

productivity”, the responses almost behave like a mirror’s reflection, meaning 40% of the

students disagree in some form with this statement and 39% agree. Only 54% of the surveyed

students agree that collaborative platforms enable them to accomplish academic tasks

quickly and only 43% agree they make it easier. However, most of the students had a neutral

opinion on both of those statements, being 4 the representative of neither agreeing or

disagreeing.

10.7. Engagement

Student 1 feels that when teachers taught classes in a similar method to the face-to-

face classes, where the teacher would make oral presentations, she was less engaged, as

there was no direct interaction between students and teachers. The most interactive classes,

with more use of tools and functionalities, captivated students' attention, and incentivized

participation. Therefore, in her opinion, collaborative platforms can help with direct

communication and in diminishing the barriers between students and professors. It is the

third time Student 1 refers to the difference she felt in the beginning, where there was the

novelty factor. She felt they were fun to use, more useful, and that she could stay more

engaged in the beginning. But with time they started to have the opposite effect on her

engagement, which she associates with habit and boredom from being connected to the

computer for too long.

88

Student 2 was not able to consistently pay attention “(…) only with the recordings. It

was what saved me”. She explains this was due to being in the comfort of her home, which

can easily lead to distractions, or when surfing the internet, where she would easily get

dispersed during online classes. Student 2 felt more engaged during more practical classes,

where the teacher would discuss clinical cases or present demonstrative videos.

Student 3 did not manage to pay attention, consistently. She did not feel motivated to

pay attention and study, and she felt her engagement got progressively worse since she felt

restless from always being trapped at home. Student 3 prefers to study with her colleagues

and feels the most motivated when she is surrounded by her colleagues. She felt her mobile

phone damaged her engagement, with the added factor of “(…) always being locked in the

same place without being able to leave, I think that was a distraction and a loss of motivation

in itself.” The classes she would feel more engaged in were the more practical classes and the

classes where she was required to have the camera turned on and the teacher would interact

and promote students’ participation.

Student 4 felt it was difficult to consistently pay attention because there were many

distractions at home, caused by his pets, family members, or neighbors. Student 4 perceives

his engagement on face-to-face classes is higher when compared to online classes. He would

feel less interested in the online classes that resembled the model used for F2F classes, which

is the teacher speaking with the aid of a PowerPoint. Student 4 would feel more engaged in

more creative tasks, such as the creation of a video, and classes that used more of the

collaborative platforms’ functionalities, such as the escape rooms, as they were a novelty to

him and therefore it would captivate his interest and attention.

Student 5 would pay little attention during his online classes and would not attend

some of his classes because he thought they were very tedious. He also acknowledges if these

classes would have been F2F the narrative would have been different, as an online learning

environment is not the most favorable for him to stay attentive. He would feel the most

engaged in more practical classes.

Student 6 felt neutral about consistently paying attention, as he felt it was easier to

get distracted in an online learning environment, but still managed to do it. When asked about

89

the tasks he enjoyed the most doing using collaborative platforms, Student 6 referred to the

oral presentations, as he felt less nervous.

Student 7 was not able to consistently pay attention during online classes, so he would

record all the sessions for himself, even if the teachers would not allow it, so he could watch

them later. He felt the most engaged when there were more interactive tasks, such as the

discussion of a theme with the screen share of programming examples.

Student 8 felt it was a bit difficult to consistently pay attention and he attributes it to

a couple of reasons. One is the size of the screen, as “(...) the size of the screen is proportional

to the attention span of the spectator (…)”. Therefore, “when we are looking at a screen

computer, and we are watching the professor in a small window, and he is not with us and all

students are in (…) other small windows, and also are not with us (…) it is hard to pay

attention”. The last reason is the lack of accountability. Since students do not feel observed

by their teachers and colleagues, they end up getting distracted on the computer, as they feel

no one is going to hold them accountable or penalize them for their behavior. The student

felt the most engaged in classes where there was a discussion among students and with the

teacher.

Table 19 - Questionnaire responses to engagement’s items

Item Indicator Mode Disagree Neutral Agree

I am able to consistently pay

attention during online classes.

ENG1 1 62% 13% 25%

I am interested in the tasks

done during online classes.

ENG2 4 36% 22% 43%

Many of the survey students were not able to consistently pay attention during online

classes, such as most of the students actually completely disagreed with the statement “I am

able to consistently pay attention during online classes”. Most of the surveyed students felt

neutral about their interest in the tasks done during online classes. Overall, 43% of the

surveyed students agreed in some way with the statement, and 36% disagreed.

90

10.8. Environment

Student 1 perceives that her home environment hindered her level of engagement,

because her entire family was confined to the same place, which was only positive for

improving her family tights, but not for her study habits and engagement during tasks. She

also found there were other interesting activities at home and that it was easier to feel too

relaxed due to the comfort of her home. Therefore, she felt she could not be fully engaged

during her academic tasks. Furthermore, her primary occupation is being a student and with

the COVID-19 pandemic she had the opportunity of going back to her hometown and living

with her parents, therefore she did not have to balance a job and domestic tasks with her

academic tasks. Her family encouraged her in her online learning, and they provided her with

a separate space for working and studying.

Student 2 felt that with online classes, she could have more time to balance all things

in her life, since she was already at home and did not have to lose time in commutes. For

online learning, Student 2 felt the environment of her home was not the best to keep her

engaged in her academic tasks. She perceived her family was neutral about their support for

her online learning, but they provided her with a favorable environment for her academic

tasks.

Student 3 felt that since she was already at home, it was quicker to do other tasks

unrelated to her course. Therefore, combining all of her responsibilities with her academic

tasks did not negatively impact her and it was actually easier because she was already at

home. Her internet connection did not cause problems, her family was supportive and

provided her with a favorable environment for online learning, since both of her parents are

teachers, and her sister is also a student, and everyone had their one separate place for

working.

Student 4 felt that combining his career with the academic tasks was more practical

because he gained time since he was already at home and did not have to commute. However,

he did not feel accomplished in his career because of the repetitive and incomplete way he

had to train his team’s players. His family supported him in his online learning, but they did

not always provide him with a favorable environment for his academic tasks, as they would

91

contribute to a distracting environment at home. Furthermore, Student 4 experienced

problems with his internet.

Student 5 experienced distractions at home such as his kitchen, the games on his

computer, and his mobile phone. He felt neutral about the impact of balancing all of his affairs

on his academic effectiveness and also about his family encouragement. However, they

provided him with a favorable quiet environment, as he has his own room, where he closes

the door and puts his headphones on.

Student 6 feels combining his academic task with other tasks at home benefited him,

as the time gained from not having to commute every day and the practicality of already being

at home, would give him more time and so, “as soon as classes would finish, I could go do the

tasks”. Student 6 did not feel there were distractions at home, however he felt he got more

easily distracted at home than in a school environment. He felt supported by his family on his

online learning, as his sister was also having online classes. Student 6 also had a favorable

environment for the fulfillment of his academic tasks.

At home, Student 7 suffered distractions that hindered his focus on his academic tasks.

For months there was construction work being done at his house, so, every day, he would be

bothered by the noise. Furthermore, he is part of a family of 6, so “...at times it would become

a bit complicated, but nothing some headphones wouldn’t resolve.” Student 7 perceives that

online learning facilitated the balancing of all his tasks, since there was more time from

already being at home and not having to commute. His family supported him, but they

perceived F2F learning as better, and they also provided him with the best environment they

could.

Student 8 perceives that when he has F2F classes, his time is more constrained

because of the commutes and the time he must stay at his university for a day of classes, so

he feels forced to better organize his time. On the other hand, when having online classes at

home, Student 8 is in an environment propitious to procrastination, so even though he had

more time to do the tasks, he would find it complicated. He felt his family was neutral about

their encouragement with his online learning, but they provide him with a favorable

environment, such as an internet connection with a larger bandwidth.

92

Table 20 - Questionnaire responses to environment’s items

Item Indicator Mode Negative Neutral Positive

At home, I have no distractions

that hinder my focus and

commitment to academic tasks

ENVIR1 2 52% 15% 34%

Combining my career, house

chores, and academic tasks have

made me academically more

effective

ENVIR2 4 42% 21% 38%

The internet connection did not

inhibit my online learning

ENVIR3 1 45% 11% 44%

The people I live with encourage

me in my online learning

ENVIR4 7 17% 24% 59%

The people I live with provide me

with a favorable environment to

accomplish my academic tasks.

ENVIR5 7 15% 15% 69%

The surveyed students felt positive about the encouragement of the people they live

with, in their online learning, and about the people they live with providing them with a

favorable environment for accomplishing their academic tasks. Actually, the most frequent

answer was “completely agree”, for both statements. On the contrary, most students

answered they completely disagreed about the internet connection not inhibiting their online

learning. However, the overall opinion was divided with 44% of the surveyed students

agreeing with the statement to some level, and 45% disagreeing to some level. When asked

about their opinion on the statement: “combining my career, house chores, and academic

tasks have made me academically more effective”, students' most frequent answer was

neither agreeing nor disagreeing. However, overall, 42% of the surveyed students disagreed

with the statement to some level and 38% agreed. When questioned about their opinion on

the statement: “at home, I have no distractions that hinder my focus and commitment to

academic tasks”, the most frequent answer was a 2 and 52% disagree with the statement in

some way.

93

10.9. Individual Performance

Student 1 thinks collaborative platforms benefited students’ performance at the time

of the pandemic, and it was the best option available. However, she feels it is not a viable

alternative to the traditional face-to-face education and it could never replace it. Student 1

perceives medical students in their clinical years, such as herself, will probably find it is harder

to learn through collaborative platforms. However, she perceives students who only have

theoretical classes will have a different opinion. Student 1 did not feel her marks improved

with collaborative platforms, as her home environment led to procrastination due to not

being able to leave her house.

Student 2 felt she learned better with collaborative platforms because of the online

recorded classes and was also able to obtain a clear understanding of the subjects taught in

class. Furthermore, she perceives collaborative platforms contributed positively to her marks.

However, when it comes to achieving the learning goals asserted for her course, as it is a

theoretical-practical course, “the practical component was missing, which is something that

the platforms do not allow to do”.

Student 3 has the perception that collaborative platforms did not contribute positively

to her individual performance. She recognizes that in a pandemic situation, where people

can’t attend face-to-face classes, collaborative platforms “(...) were the best solution we

found, and it is a lot better than nothing (…)”. However, Student 3 feels her education “(...)

was very incomplete”, as “(...) there were a lot of things I should have learned, and I did not

(…)”. This is because she feels her course is incompatible with distance education, as it

requires students to learn through in-person contact with patients and through practice.

Furthermore, Student 3 does not feel she learns better with collaborative platforms, as

practical classes were not recorded, so her individual performance ended up being negatively

impacted. In terms of theoretical classes, Student 3 feels there is a more positive outcome, as

the sessions were recorded, and even if she did not pay attention or attended classes, she

would still be able to benefit from the recordings. This method was also more compatible

with the way she likes to work, as before the pandemic, Student 3 would not attend practical

classes and would study through “desgravadas”. Student 3 felt she did not gain a clear

understanding of the subjects taught in class. However, she perceives she managed to achieve

the learning goals asserted for her course’s theoretical part, as she managed to attain good

94

grades in the written exams. In terms of clinical thinking and practical classes, she feels she

was really affected, as there were no F2F classes and in turn, her marks in those subjects were

worse than what they could have been.

Student 4 perceives collaborative platforms had a positive impact on his performance,

when it comes to group projects, and in an online context he learns and works better with

them. Furthermore, he feels they contributed positively to his marks. However, in comparison

to F2F classes, he feels he learns better in-person. Student 4 perceives he had a clear

understanding of what was taught in class and that he managed to achieve the learning goals

asserted for his course, as he managed to graduate. However, his course “(...) has a practical

part that felt short due to the pandemic (…)”, but he managed to accept that.

Student 5 perceives collaborative platforms did not have a positive impact on his

performance, but he feels they were crucial for the continuance of classes during the COVID-

19 pandemic. He also feels that in an online learning context, collaborative platforms help

students learn better. However, compared to F2F, F2F “(...) is a lot better, it’s on a different

level”. Student 5 perceives he did not achieve the learning goals asserted for his course, and

some of his marks were better and others worse.

Student 6 perceives collaborative platforms had a positive impact on his performance

and they helped students adapt to a new reality. He also perceives his marks improved with

online learning. In his opinion: “I think I probably learn better online. Combining everything, I

think the best is online learning with some (F2F) practical classes but combining everything".

Furthermore, Student 6 felt he learned everything he would have learned in F2F classes, and

he managed to achieve the learning goals asserted for his course.

Student 7 felt collaborative platforms contributed to a better performance, as they

allowed him to posteriorly watch the online classes, through recordings. Also, they helped

dynamize the making of group projects. In an online learning environment, Student 7

perceives he learns better with collaborative platforms, as he had teachers using pre-

recorded classes and he prefers to have live classes, since he can ask questions directly and it

is closer to an in-person interaction. However, he perceives F2F classes are the best way to

have classes. Furthermore, Student 7 felt he managed to attain the learning goals asserted

95

for his course and he feels his marks were slightly better, but he attributes that to working

harder.

Student 8 perceives collaborative platforms certainly had a positive impact on his

performance. Even though he recognizes students are more engaged in a F2F learning

environment, when learning online, Student 8 would “(...) research more, read more, and

study more (…)”. Therefore, he felt he learned more when he was in an online learning

regime. Furthermore, Student 8 thinks that “(...) it wasn't because the classes started to be

held on Zoom that it became more complicated to understand”, so he feels he gained a clear

understanding of the subjects taught in class. Also, Student 8 perceives he managed to

achieve the learning goals asserted for his course. In terms of his marks, Student 8 feels all

collaborative platforms had a positive contribution, he even finds them essential for his

cinema course.

Table 21 - Questionnaire responses to individual performance’s items

Item Indicator Mode Negative Neutral Positive

Collaborative platforms have a

large positive impact on my

performance as a student.

PERF1 5 31% 19% 51%

I learn better with collaborative

platforms than without them.

PERF2 4 41% 22% 38%

With the help of collaborative

platforms, I have gained a clear

understanding of the subjects

taught in class

PERF3 5 35% 20% 45%

With the help of collaborative

platforms, I managed to achieve

the learning goals asserted for my

course

PERF4 4 33% 22% 45%

Collaborative platforms positively

contributed to my grades.

PERF5 5 35% 17% 48%

96

51% of the surveyed students had a positive perception about the collaborative

platforms’ positive impact on their performance as students, and the most reported answer

was “somewhat agree”. Most students had a neutral perception about learning better with

collaborative platforms. However, 41% disagreed in some way with the statement: “I learn

better with collaborative platforms than without them”; and 38% agreed. Making it the only

statement with more negative perceptions than positive. Continuing with our analysis, we can

assess that 45% of the surveyed students had a positive perception about the statement:

“With the help of collaborative platforms, I managed to achieve the learning goals asserted

for my course”, and 35% felt negative about it. The most frequent answer, when asked to

classify that statement, was somewhat agree. Most students had a neutral perception about

achieving the learning goals asserted for their course, with the help of collaborative platforms.

However, 45% agreed in some form with that statement and 33% disagreed. Lastly, 48% of

the surveyed students agreed in some way with the statement: “Collaborative platforms

positively contributed to my grades”; and 35% disagreed, and the most frequent answer was

“somewhat agree”.

11. Discussion

Valuable answers need to be aesthetic, by arising from the powerful simplicity of the

answer, scholarly, by advancing the area under study in fundamental ways that influence

future progress, and of practical utility by having the potential to make the practical world

better in important ways (Rai, 2017). Therefore, for the discussion of our results, to provide

better and more accurate inferences (Venkatesh et al., 2013), we are going to combine and

explain our quantitative results, with the support of our qualitative results and knowledge

base, to form a valid and relevant theory.

The Task-Technology Fit theory was originally developed to explain how the fit

between the tasks performed by an individual and the technology used to perform those tasks

would impact individual performance, in an organizational setting. It is interesting to

understand how it can be extended to other contexts, such as the context of online learning

in higher education.

97

Against what was originally theorized by Goodhue and Thompson, we found the

relationship between USE and PERF to be nonsignificant. Moreover, we found new and more

significant relationships and confirmed the relationship between TTF and PERF.

Before we start the results’ discussion, we need to do a statistical power analysis, so

we can make inferences about the relationships between variables in our model, confidently.

This is especially important because, in our model, the relationship between USE and PERF

was not significant and this goes against the status quo in the literature. Therefore, we need

to determine if our model is powerful or strong enough to detect the significant causal effects

on the variables, so we can confidently verify or deny our hypothesis.

To calculate our post-hoc statistical power, we used the software “Post-hoc Statistical

Power Calculator for Multiple Regression” (Soper, 2022). For the calculation, we needed the

number of performance’s predictors in the model (6), the observed 𝑅2(0.731), the probability

level or significance level (0.05), and the sample size (295). We managed to obtain an

observed statistical power of 1.0, which means there is a 100% chance of finding statistically

significant effects. Likewise, using the software G*Power version 3.1.9.7 (Faul et al., 2009,

2020), our post hoc achieved power, to detect an effect size of 0.15 and 0.02, was 1 and 0.80,

respectively. So, if there was a significant effect of USE on PERF, we were bound to find it.

Therefore, we are confident in our inferences about the model (Goodhue et al., 2006).

11.1. Technology Characteristics and Task Characteristics

Overall, the 8 interviewed students regard collaborative platforms’ functionalities as

useful, as the surveyed students do. The two most useful functionalities for the interviewed

students are screen sharing and meeting recording, followed by the chat and the breakout

rooms. It was also mentioned the polling feature, the raise hand button, Zoom’s meeting

participants' large capacity, and TeamViewer’s remote desktop feature. All students used

collaborative platforms, especially Zoom, for group projects, theoretical and practical classes,

and oral presentations.

In our qualitative study, similarly to the survey’s responses, students experienced

some form of unexpected problems during their experience with collaborative platforms.

Those problems were related to the internet connection, the Zoom meeting maximum

capacity, Zoom meeting maximum duration, and the realization of the final exams.

98

Students had to collaborate with others for tasks like group projects, but also online

classes that used the breakout rooms feature from Zoom, and other more interactive classes.

The 8 interviewed students also recognized the importance and necessity of real-time

communication facilitated by platforms like Zoom, because they provide a viable alternative

to face-to-face communication and avoid place constraints. However, the interviewed

students also recognize that they can’t provide the same experience as F2F classes, which is

very accentuated in practical classes, due to the possibilities of collaborative platforms.

The existence of a relationship between TASK and TECH and TTF is consistent with the

findings of previous studies (Alazab et al., 2021; Dishaw & Strong, 1999; Goodhue &

Thompson, 1995; Staples & Seddon, 2004; Virdyananto et al., 2016). TECH and TASK have a

positive effect on TTF, but there is a significant discrepancy between the size of the effect of

TASK and TECH on TTF. TECH has a strong effect on TTF, whereas TASK has a weak effect on

TTF. Thus, removing the path between TASK and TTF, would cause a very little decrease in the

𝑅2 value of TTF but removing TECH would cause a significant decrease. So, we can state that

a high level of TECH leads to a significant increase in TTF, and a high level of TASK leads to a

small increase in TTF. Similarly, in the context of distance education, Raven et al. (2010), Yu

and Yu (2010), and Wan et al. (2020) found that technology Characteristics had a significantly

larger impact on TTF.

This is because, during the COVID-19 pandemic, students were extremely dependent

on the functionalities and characteristics collaborative platforms offered, to perform their

tasks. So much so that some tasks had to be adapted to fit collaborative platforms'

functionalities and characteristics. An example of that is how practical classes had to be

adapted or canceled so they could fit into the possibilities of these platforms. Also, daily

classes were very dependent on the lack of technical problems and on functionalities that

allowed professors to interact with their students. Furthermore, for realizing group projects

and presenting them, students were completely dependent on real-time distant

communication. Thus, collaborative tasks that require real-time communication, and

frequent coordination with the professors’ and students’ efforts, are very dependent on

collaborative platforms' functionalities and quality.

In conclusion, the quality of collaborative platforms, their capacity to provide real-time

communication, having no frequent and unexpected problems, and having utilitarian

99

functionalities is going to contribute greatly to predict how suitable collaborative platforms'

functionalities are to perform academic tasks, if the functionalities feel enough to students,

and if they can meet all aspects of students’ learning requirements.

11.2. Task-Technology Fit

Overall, most of the surveyed students had a high level of task characteristics and

technology characteristics, which naturally led to a high level of TTF. The same tendency

applies to the interviewed students. This means, for both samples, collaborative platforms

provided an adequate set of functionalities for students to perform their academic tasks. It is

also interesting to note that, in both survey responses and interviews, when it comes to

learning requirements, there was a more divided perspective. This happens because students

felt that collaborative platforms were adequate to complete the academic tasks proposed by

the professors. However, courses with a more practical nature had to make serious

adaptations to practical tasks, as most had to be canceled, reduced in quantity, or substituted

by less effective alternatives. Thus, students enrolled in courses with a stronger practical

component felt that all their learning requirements were not met by these platforms.

However, students enrolled in a more theoretical course did not feel a significant difference,

and thus experienced a higher level of TTF.

TTF affects USE but the size of the effect is small, which means if we removed this

relationship, there would be a small decrease in the 𝑅2value of Utilization. So, we can only

say that an increase in TTF will lead to a small increase in Utilization. Our findings are in

accordance with the findings of Goodhue and Thompson (1995) and Dishaw and Strong

(1999). However, in the context of education, Gerhart et al. (2015) found a strong path

between the two. This can be explained, due to the nature of the technology and

voluntariness of USE. For instance, in that study, students could USE e-textbooks for learning

if they wanted to (USE was voluntary), so if they felt e-textbooks were fit for learning, then

they would use them with increased frequency, if not they would not. That is why this was

the strongest path in their model. However, in the context of online learning, students are

required to use the learning technology whether they find there is a low or high TTF. This also

supports the important role of VLNT as a moderator of the impact of TTF on USE.

100

TTF has a statistically significant impact on PERF, which is in accordance with the

findings from McGill and Klobas (2009), Gerhart et al. (2015), Harrati et al. (2017), Bere (2018),

and Jardina et al. (2021), in the context of distance education. However, its contribution to

explaining PERF is almost irrelevant. In fact, removing this path from our model would cause

a minimal impact on the 𝑅2value of PERF. When combining the results from the quantitative

and qualitative study, we can infer students consider TTF to be important. However, it looks

like it is more relevant in future voluntary use, where students are able to choose which tasks

they can perform using collaborative platforms. This is highlighted by the fact that 7 out of

the 8 interviewed students would like to continue using these platforms for blended learning,

where they could use collaborative platforms for tasks they perceived had a higher Fit with

the technology, and thus impacted their performance positively. Furthermore, there are

other factors they perceive to have had a bigger impact on their Individual Performance as

online learning students.

11.3. Facilitating Conditions

For all the interviewed students there was a high level of Facilitating Conditions. They

all had the necessary resources, meaning they had a computer and connection to the internet.

Therefore, they were able to attend classes online and work with their colleagues via

collaborative platforms. Furthermore, all 8 students felt very positive about their knowledge

in using collaborative platforms, as they considered them easy to use tools with a short period

of adaptation, except for one student. Additionally, if difficulties arose, there was help readily

available to all students. Similarly, the surveyed students had a high level of FC.

Facilitating Conditions (FC) plays an important role as a precursor of utilization, this

role had already been theorized by Goodhue and Thompson (1995), but they never tested it

in their study. The impact of FC on USE, when compared to the impact of TTF on USE, is much

greater. In fact, the 𝑓2, is only strong for the relationship between FC and USE, so removing

this relationship is going to create a significant decrease in the 𝑅2value of USE, whereas

removing the relationship between TTF and USE would create a minimal decrease, as we had

established before. Our findings are consistent with previous studies in the context of

distance education, where FC was the strongest predictor of USE (Khechine et al., 2020;

Masadeh et al., 2016; Yi et al., 2016; Zainol et al., 2017).

101

Facilitating Conditions had a very high level among our respondents in both the

quantitative and qualitative study, as well as the level of USE. If there are factors that facilitate

the use of a technology, such as having the resources, knowledge, and technical support,

students will feel more comfortable in actively using collaborative platforms, and thus FC ends

up being a much bigger predictor of USE. Therefore, the students’ ability and frequency to

use collaborative platforms is favored by high Facilitating Conditions.

Whereas TTF’s impact on USE is conditioned and diluted by mandatory USE. As

students are required to use technology for online learning, a task with a high fit, is not going

to contribute much more than a task with a low fit, for increasing the level of USE. When USE

is voluntary students perceive there is a much higher TTF, as they can choose what tasks they

want to perform utilizing collaborative platforms. Therefore, they will be much more inclined

to use these platforms, as they will only use them voluntarily for tasks they perceive to have

a high TTF.

11.4 Utilization

In our study, Utilization does not affect performance, which goes in line with the

findings of Pentland (1989) and Staples and Seddon (2004). In fact, this relationship has the

lowest effect size (i.e., .001), in our model. So, this means removing this path from our model

would have no impact on the 𝑅2value of Individual Performance. This is due to several

reasons.

According to Goodhue and Thompson (1995), when USE is mandatory, performance

is going to be increasingly impacted by TTF, which was the case for our study. Considering

there was a high level of VLNT, students felt that USE was not dependent on their criteria,

therefore, even if they frequently used collaborative platforms, and extensively used their

functionalities, they did so because they felt their teachers required them to do so. They were

also obliged to do so due to the pandemic context, which did not allow them to meet in person

with their colleagues and professors. Thus, they did not really have other options besides

using these platforms for online learning and for communicating, interacting, sharing, and

collaborating.

Pentland (1989) found that even though IRS auditors had a positive attitude towards

the use of Personal Computers and used them extensively, utilization had few positive

102

impacts and possible negative impacts on performance, as using an IT with a low TTF can

actually hurt productivity. Thus, only the agents appropriately selecting tasks for computer

use would achieve an overall productivity gain. This extends the notion presented before, as

students were not able to select the most fit tasks for collaborative platforms and they had

to do everything using these platforms. It is only natural that USE ends up not being a

predictor of PERF.

Another possibility, already approached by Tam and Oliveira (2016), is the participants

age. For younger people, USE is not a relevant predictor of PERF, as they already associate

technology to performance (Tam & Oliveira, 2016a). This generation of university students is

more technologically savvy than any previous generation, and they are proficient in the use

of social communication tools such as Facebook and Instagram, and most students have a

mobile phone and laptop (Kalin, 2012). That is why they are called the Net Generation or

Digital Natives (Kalin, 2012). Being so, students give more importance to TTF and other

factors, and assume it has no impact on their individual performance (Tam & Oliveira, 2016a),

as it was mentioned by some of the interviewed students.

Finally, the novelty factor (Venkatesh et al., 2012) was referred by 2 students during

the interviews. In the beginning they felt more attracted to utilizing collaborative platforms

and would stay more engaged due to the curiosity and excitement of discovering and

experiencing something new. However, the increase in USE started to have the opposite

effect, or a negative effect, due to the tiredness and boredom from being connected to the

computer for too long. Therefore, as students were using these platforms every day for

several hours, USE started to have the opposite effect on their PERF.

After students experienced collaborative platforms’ effects on their individual

performance, they were able to assess if their expectations were met or not, and that

influenced future USE (Goodhue & Thompson, 1995). 7 out of the 8 interviewed students,

when questioned about their future utilization, answered they would like to continue their

course in a blended learning regime. This is in accordance with the findings in Flores et al. 's

(2021) investigation, where the questioned university students, in a future setting, would also

prefer blended learning.

103

This is because, as it is described by Goodhue and Thompson (1995), when students

utilized collaborative platforms, they were able to experience how they affected their

performance, and from that they created a feedback. Students felt collaborative platforms

met and even exceeded their expectation for theoretical classes, for attending meetings, for

doing group projects, and collaborating with their colleagues. Having recorded lectures, the

convenience and comfort of having classes at home, the avoidance of commutes, and better

time management, were all things that contributed to that positive perception.

However, when it comes to practical classes, students were underwhelmed and felt

prejudicated by the USE of collaborative platforms. So, after their experience, for practical

tasks, they expected negative consequences of USE on their PERF. Therefore, they felt F2F

classes were the most advantageous method for practical tasks. In terms of future use, this

could change students' expected consequences of utilization and allow them to improve their

individual technology fit and the overall TTF.

11.5. Perceived Usefulness

The 8 interviewed students can perceive collaborative platforms’ advantages and

usefulness, as it seems to be the case for the surveyed students. However, the interviewed

students, with a more practical course, can feel more conflicted when it comes to its

usefulness, as they are not useful for practical classes. Nevertheless, 75% of the interviewed

students considered collaborative platforms very useful and advantageous for group projects

and to collaborate with other students, 63% considered them useful tools for learning, and

50% mentioned they were useful for practical classes and studying, as screen recordings allow

students to rewatch the lectures and study the contents more in depth. 50% also mentioned

they were useful for real-time distance communication, 38% mentioned they were useful as

they removed the necessity for commuting, and 25% mentioned they allowed them to

organize and save time. When it comes to productivity, the surveyed students had a more

negative perception but somewhat balanced between negative, neutral, and positive. The

same happened for the interviewed students, as 2 students considered collaborative

platforms contributed positively to their productivity, 3 students were neutral in that regard,

and 3 students felt they did not increase their productivity.

104

Considering these factors presented by the interviewed students, it is clear why

Perceived Usefulness had a positive and the most significant impact on Individual

Performance. Students who perceive collaborative platforms as advantageous and useful, will

be able to perceive better impacts on their performance and will likely identify them as

meaningful tools for positive outcomes.

The impact of PU on PERF is also confirmed by previous studies (Ali & Younes, 2013;

Bravo et al., 2015; Raven et al., 2010).

Furthermore, when students perceive collaborative platforms as advantageous and

useful, they will perceive them as more compatible with their needs, work style, course, and

situation, as one of the differentiating factors that contributed to the different perspectives

on PU was the dichotomy of practical classes vs. theoretical classes, and students enrolled in

courses that traditionally need more F2F contact with people, tend to feel collaborative

platforms are not as compatible with them.

11.6. Compatibility

As in the survey responses, the level of COMP for the interviewed students was lower.

Only 3 out of the 8 interviewed students felt collaborative platforms were compatible with all

aspects of their course. It is interesting to note that these 3 students are enrolled in courses

with a strong theoretical component and a practical component that does not require physical

contact with others or physical movement. Compared to medicine and sports sciences,

engineering, cinema, and geography are lesser practical courses, that do not require F2F

interactions with other people, and therefore these students did not feel such a negative

impact from the adaptation of F2F classes to an online environment.

Soria et al. (2020) similarly to our findings, found contrasting results when studying

students' perception on the appropriateness of course content for remote learning. For

example, architecture, visual and performing arts, and education students, comparatively to

other students enrolled in different courses, found their course content was less appropriate

for remote learning.

When it comes to compatibility with the way students like to work, for the surveyed

students, the perspectives were considerably more negative. This can be explained by

students feeling the need to have human interactions and preferring to work F2F with people.

105

A factor that was mentioned by 6 out of the 8 students. However, students understand that,

considering the pandemic situation, it was a viable option and the best solution found at the

time.

Even though most students had a more negative perspective when it comes to

Compatibility with collaborative platforms, they were still able to identify situations where

they felt this was possible. For instance, compatibility with attending theoretical classes

seems to be connected to students that found screen recording an advantageous feature for

studying. Also, students that do not live in Lisbon prefer to attend meetings remotely, as it

avoids commutes and saves time. It seems clear why when there is a higher level of PU there

is going to be a higher level of COMP.

As an example, the student enrolled in a Geography degree found collaborative

platforms very useful learning tools and a more efficient and advantageous way for delivering

education and learning. The same students also had a very high level of COMP, as he felt

collaborative platforms were compatible with all aspects of his course, the way he likes to

work, as he can be more efficient, and feel more comfortable doing oral presentations. These

further highlights how someone viewing collaborative platforms as advantageous and useful,

is thighed up with their fit with a technology, as an individual.

Therefore, PU seems to have an added impact on PERF when it impacts COMP, as

students who find collaborative platforms useful and compatible will have an even greater

level of PERF, and students who find collaborative platforms useless and incompatible will

have an even lower level of PERF.

11.7. Environment

For the interviewed and surveyed students there was a tendency to feel they had an

adequate space and a supportive family. Students felt that balancing life and other

responsibilities can be easier as they had more time. However, having other things to do at

home can be a distraction, as it can lead to procrastination, when it comes to performing

academic tasks.

Students feel that an adequate environment at home is important for their

engagement, as a distracting learning environment makes staying engaged harder. Also, a

good internet connection seems to be of great importance to students as it was recognized

106

by the interviewed students, the effective use of these platforms is dependent on the

internet. Although the environment did not have a direct impact on PERF it had an indirect

impact through ENG. This means that only when a favorable environment or an unfavorable

environment impacts students ENG positively or negatively, students are going to feel an

impact of the ENV on their PERF.

11.8. Engagement

The surveyed and interviewed students had an overall lower level of ENG, especially

when it comes to consistently paying attention, which refers to behavioral engagement.

However, students were able to feel more emotionally engaged in the tasks done during

online classes. This was the same case for 7 of the 8 interviewed students, who were not able

to consistently pay attention, whereas 1 student felt neutral. It is interesting to note that the

interviewed students automatically identified the impact of their home learning environment

on their level of engagement. Furthermore, the interviewed students felt the most interested

and engaged in classes that went far from the traditional model, identified as the teacher

speaking with the support of a PowerPoint presentation. They preferred tasks that generated

interactions among students and with the professor, as it made them feel more included,

tasks that generated discussions, and tasks that used more of collaborative platforms’

functionalities, which were innovative to them.

Accordingly, Soria et al. (2020), concluded that one of the biggest obstacles among

university students was the lack of interaction or communication with other students. A factor

that was cited by the interviewed students.

We can conclude from our interviews that most students felt more engaged, when

using new functionalities from Zoom, or when new teaching methods were introduced

through more interactive and creative tasks. This can contribute to making students more

consistently engaged and motivated and contribute to making classes feel less monotonous.

Similarly, Croxton (2014) concluded that to promote achievement of learning

outcomes, satisfaction, and persistence in online courses, there needs to be a social online

learning environment that promotes active student engagement with course materials and

meaningful and purposeful interactions with teachers and colleagues. Students’ engagement

has been proven to increase with collaboration during the learning process, which also

107

enhances co-creation, satisfaction, and commitment to the online course Francescucci and

Rohani (2019), thus affecting student’s performance.

11.9. Individual Performance

Overall students had a more neutral to somewhat positive perspective of the impact

of collaborative platforms on their PERF. What negatively impacted and thus weighed down

this perception was, especially, the level of ENG, which has a large effect size on PERF, and

thus students having an overall lower level of engagement ended up contributing negatively

and significantly to their perception of PERF. Also, having a distracting ENV with a bad internet

connection heightened the more negative level of ENG and thus contributed negatively to

students’ PERF.

Compatibility also contributed negatively to students’ level of PERF. However, as its

effect size is small the impact on PERF is not going to be as accentuated. Nevertheless, COMP

has an important mediating role, as PU, the strongest predictor of PERF, has an even stronger

impact when influenced by students' level of COMP. Overall, for PU, students had a neutral

to slightly positive perspective, and that was reflected in students' perception of PERF,

especially when heightened by the levels of COMP.

12. Implications

12.1. Theoretical Implications

Facilitating Conditions had not been tested as a moderator of TASK and TECH’s impact

on TTF. However, we were only able to detect a statistically significant moderating role of FC

on TECH. This probably happened due to the concepts’ nature, as FC relates to the knowledge

and resources necessary to use collaborative platforms, and technical support, and thus it is

more related to the characteristics of a technology. Therefore, a higher level of FC ends up

influencing positively how much students perceive a technology is functional, high quality,

and utilitarian.

108

Graphic 1 - Moderating Effect of FC on the relationship between TECH and TTF

Graphic 1 shows us a two-way interaction between TECH and TTF, moderated by a

third variable, which is FC.

We can understand, as can be seen in Graphic 1, that when there is a low level of TECH,

the impact of FC is very similar. Although a high level of FC still leads to a higher level of TTF

on students with a lower level of TECH (2.36 vs. 2.46).

When there is a high level of TECH, the impact of FC is more accentuated. And as

expected, a high level of FC leads to a higher level of TTF (3.41 vs. 3.78).

Therefore, we can conclude that the higher the level of FC is, the stronger will be the

impact of TECH on TTF. Plus, the higher the level of TECH, the more impactful the role of FC

as a moderator is going to be.

As we expected, the role of VLNT as a moderator of the effect of TTF on USE, is very

significant, and helped us explain not only this relationship, but also USE’s impact on PERF

and confirm the idea of PERF being increasingly impacted by TTF when USE is not voluntary

(Goodhue & Thompson, 1995).

109

Graphic 2 - Moderating Effect of VLNT on the relationship between TTF and USE

Graphic 2 shows us a two-way interaction between TTF and USE, moderated by a third

variable, which is VLNT.

We can understand, as can be seen in Graphic 2, that when use is mandatory, whether

there is a low or high TTF, the impact on USE is going to be practically the same (3.2 vs. 3.25).

However, when the USE is voluntary and students perceive there is a higher TTF, there is going

to be a higher level of USE. On the other hand, in a voluntary setting, when students perceive

there is a lower TTF, their level of USE is going to be significantly lower (2.47 vs. 3.08). Also, it

is interesting to note that the level of USE is higher in a mandatory context.

This highlights the fact that students under a circumstance of mandatory USE, were

required to use collaborative platforms frequently and for a great number of tasks, more than

students under a voluntary use. However, this has a major downside, as even when there was

a low TTF, students were obligated to USE collaborative platforms. That is why, with

mandatory USE, TTF has a much more important role as a predictor of PERF.

The graph also underlines how under a circumstance of voluntariness TTF can impact

USE more significantly, as students who perceive a lower TTF will opt to USE collaborative

platforms less, since they know it will impact their performance negatively. However,

110

students who perceive a high TTF will USE collaborative platforms almost as much as students

who are required to USE them, as they know there will be a positive outcome.

12.2. Management Information Systems (MIS) Implications

This study provides findings that are useful for the developers of learning technology,

higher education institutions, teachers, and researchers who want to study the impact of

information technology on students’ individual performance.

We were able to find gaps in the literature when reviewing the previous research on

collaborative platforms, in the context of technology use in higher education. There have

been very few attempts to advance a theory that explains how collaborative platforms can

affect students' individual performance. Furthermore, the published research focuses mainly

on adoption and acceptance of collaborative platforms. For these reasons, our study extends

the body of knowledge in Information Management Systems.

The theory developed from this exploratory research should not be exclusive to

collaborative platforms and higher education Portuguese students, since our model is flexible

enough to be adapted to other technologies and contexts in distance and blended education.

Therefore, we encourage researchers in this field to test our model. Focusing on performance

and developing a theory to understand this matter, will help us attain an understanding of

what actually predicts students' individual performance in an online learning environment.

Live video-conferencing with sound and image, and break out rooms are great

features that promote students’ active participation and capture their attention. The

developers of collaborative learning technology should focus on functionalities that promote

students’ engagement in an online learning environment, which is prone to having several

distractions. A good direction for that is developing features that promote and facilitate

interactions among students and with the professors, and innovative and creative features

that capture the attention of students and facilitate collaboration.

Teachers and educational institutions should be encouraged to use collaborative

technology not just in uncertain environments. The use of collaborative platforms can benefit

university students’ performance, but only if used under the right circumstances. They should

only be used for executing tasks adequate to the functionalities and characteristics of these

111

technologies. For subjects with a more theoretical nature, for practical exercises based on

discussion, and for group projects, collaborative technologies can be a useful and compatible

asset. Furthermore, teachers should take advantage of innovative features to capture

students’ attention and create opportunities of interaction.

Lastly, the traditional method of teaching classes seems to not work as well in Zoom.

If professors give lectures the traditional way, students tend to disperse more easily and

become more distracted by the environment, and so they feel dependent on class recordings

to study and to catch on what was missed during the live class. However, when professors

adapt their teaching method, to a more interactive method, that incentivizes students to have

their cameras turned on, promotes interaction among students and with themselves,

students feel more connected and included, and thus have a better performance during

classes.

12.3. Limitations and Future Research

This study has limitations that should be taken into consideration when generalizing

its findings. First, this mixed methods study was conducted in Lisbon, which is the capital of

Portugal, and even though the out-of-sample predictive power of our model was tested using

𝑃𝐿𝑆𝑝𝑟𝑒𝑑𝑖𝑐𝑡, to enhance generalization, future research should test our model with different

samples from different locations.

Furthermore, this study is set in the context of online higher education, but there are

other levels of online education, such as secondary education, and other types of online

education, thus it would be interesting to test how well our model would predict performance

in those contexts. Similarly, it would be interesting to test this model with other types of

technology used in the context of education.

The implication of voluntariness vs. mandatoriness should be further explored and

taken into consideration in future research, especially as a moderator of the relationship

between Task-Technology Fit and Utilization. Similarly, Facilitating Conditions seems to have

a moderating role on the relationship between Technology Characteristics and TTF, so there

should be further investigation on this relationship.

112

Further research should be conducted on the two mediation models, as these were

the strongest predictors of Individual Performance. Also, there should be a focus on

developing a scale for measuring engagement.

Finally, it would also be interesting to study how human interactions influence

engagement since it is a topic that has been approached in literature and that was approached

by the interviewed students.

13. Conclusion

The emergency use of online learning during the pandemic, brought attention to the

topic of its impact on students' learning outcomes. However, the topic of individual

performance, in the context of technology in education, has received limited attention.

Our research develops a model capable of predicting students’ individual

performance, by adding two mediation models, with two new constructs each: environment

and engagement, and perceived usefulness and compatibility, to the TTF theory (Goodhue &

Thompson, 1995). These models proved to be very relevant for predicting students’ individual

performance.

Furthermore, voluntariness has a very important moderating role, and it should not

be neglected. We are the first to test the impact of voluntariness in the relationship between

TTF and USE, and it revealed to be very important to explain how a mandatory setting vs. a

voluntary setting can have very different impacts.

Even though the impact of USE on PERF was not statistically significant, it was very

relevant to further explain the complexity of this variable, which is dependent on several

factors, such as voluntariness, TTF, age, novelty, and future use.

We found that students, after utilizing collaborative platforms, can see its benefits for

academic purposes, such that most of the interviewed students would prefer to continue

their academic path in a blended learning regime, where there would be the possibility of

online learning for theoretical classes, and F2F classes for practical classes.

For there to be a better outcome from collaborative platforms, interactions between

students and professors, and among peers should be prompted, so that students can feel

more involved. For the same purpose, in an online exclusive course, the use of the platforms’

113

functionalities and an adaptation in learning delivery should be done, as the traditional

methods used in F2F learning seem to not be as effective in online learning.

If the professor or the learning institution intends to follow a blended learning regime,

with theoretical classes via online learning, then the recording of online sessions seems to be

of great importance for students, as due to the nature of online learning, there tends to be a

loss in focus. Thus, it is important for students to revise and study from those recordings. In

fact, this was collaborative platforms’ most useful functionality for students, as it offered

them great learning support, and even better learning outcomes.

Finally, these results prove how important it is to conduct mixed methods research

and follow a rigorous methodology that gives equal importance to both the quantitative and

qualitative strands of the investigation. In that way, it is possible to produce better inferences,

corroborate our findings, and provide meaningful answers that contribute to the body of

knowledge.

114

References

Adeniyi, C. O., Oladele, E. O., & Adeniyi, E. O. (2021). Influence of Home and Career on the

Academic Performance of Married Distance Learners in South-West, Nigeria.

Pakistan Journal Of Distance And Online Learning, 6(2), Article 2.

http://journal.aiou.edu.pk/journal1/index.php/PJDOL/article/view/745

Agarwal, R., & Prasad, J. (1997). The Role of Innovation Characteristics and Perceived

Voluntariness in the Acceptance of Information Technologies. Decision Sciences,

28(3), 557–582. https://doi.org/10.1111/j.1540-5915.1997.tb01322.x

Alamri, M. M., Almaiah, M. A., & Al-Rahmi, W. M. (2020). The Role of Compatibility and

Task-Technology Fit (TTF): On Social Networking Applications (SNAs) Usage as

Sustainability in Higher Education. IEEE Access, 8, 161668–161681.

https://doi.org/10.1109/ACCESS.2020.3021944

Alanazi, A. A. (2019). Online Learning Environments: Investigating the Factors Influencing

Social Presence. https://kuscholarworks.ku.edu/handle/1808/30232

Alanazi, A. A., Frey, B. B., Niileksela, C., Lee, S. W., Nong, A., & Alharbi, F. (2020). The

Role of Task Value and Technology Satisfaction in Student Performance in

Graduate-Level Online Courses. TechTrends, 64(6), 922–930.

https://doi.org/10.1007/s11528-020-00501-8

Alanazi, A. A., Niileksela, C., Lee, S., Frey, B., & Nong, A. (2019). A Predictive Study of

Learners’ Perceived Performance in Higher Education Online Learning

Environments.

Alazab, M., Alhyari, S., Awajan, A., & Abdallah, A. B. (2021). Blockchain technology in

supply chain management: An empirical study of the factors affecting user

adoption/acceptance. Cluster Computing, 24(1), 83–101.

https://doi.org/10.1007/s10586-020-03200-4

115

Ali, B. M., & Younes, B. (2013). The Impact of Information Systems on user Performance: An

Exploratory Study. Journal of Knowledge Management, Economics and Information

Technology, 3(2), 1–10.

Al-Salman, S., & Haider, A. S. (2021). Jordanian University Students’ Views on Emergency

Online Learning During COVID-19. Online Learning, 25(1).

https://doi.org/10.24059/olj.v25i1.2470

Axelson, R. D., & Flick, A. (2010). Defining Student Engagement. Change: The Magazine of

Higher Learning, 43(1), 38–43. https://doi.org/10.1080/00091383.2011.533096

Ayala, N. F., Le Dain, M. A., Merminod, V., Gzara, L., Enrique, D. V., & Frank, A. G. (2020).

The contribution of IT-leveraging capability for collaborative product development

with suppliers. The Journal of Strategic Information Systems, 29(3), 101633.

https://doi.org/10.1016/j.jsis.2020.101633

Baykal, G. E., Van Mechelen, M., & Eriksson, E. (2020). Collaborative Technologies for

Children with Special Needs: A Systematic Literature Review. In Proceedings of the

2020 CHI Conference on Human Factors in Computing Systems (pp. 1–13).

Association for Computing Machinery. https://doi.org/10.1145/3313831.3376291

Becker, J. M., Ringle, C. M., & Sarstedt, M. (2018). Estimating moderating effects in PLS-

SEM and PLSc-SEM: Interaction term generation* data treatment. Journal of Applied

Structural Equation Modeling, 2(2), 1–21.

Bélanger, F., & Allport, C. D. (2008). Collaborative technologies in knowledge telework: An

exploratory study. Information Systems Journal, 18(1), 101–121.

https://doi.org/10.1111/j.1365-2575.2007.00252.x

Bere, A. (2018). Applying an Extended Task-Technology Fit for Establishing Determinants of

Mobile Learning: An Instant Messaging Initiative. 29, 16.

Bîzoi, M., Ana-Maria, S., & Filip, f g. (2009). Using Collaborative Platforms for Decision

Support.

116

Bollen, K., & Lennox, R. (1991). Conventional wisdom on measurement: A structural

equation perspective. Psychological Bulletin, 110(2), 305–314.

https://doi.org/10.1037/0033-2909.110.2.305

Bravo, E. R., Santana, M., & Rodon, J. (2015). Information systems and performance: The

role of technology, the task and the individual. Behaviour & Information Technology,

34(3), 247–260. https://doi.org/10.1080/0144929X.2014.934287

Brislin, R. W. (1970). Back-Translation for Cross-Cultural Research. Journal of Cross-

Cultural Psychology, 1(3), 185–216. https://doi.org/10.1177/135910457000100301

Brodahl, C., & Hansen, N. K. (2014). Education Students’ Use of Collaborative Writing Tools

in Collectively Reflective Essay Papers. Journal of Information Technology

Education: Research, 13, 91–120.

Carrillo, C., & Flores, M. A. (2020). COVID-19 and teacher education: A literature review of

online teaching and learning practices. European Journal of Teacher Education,

43(4), 466–487. https://doi.org/10.1080/02619768.2020.1821184

Castillo-Montoya, M. (2016). Preparing for Interview Research: The Interview Protocol

Refinement Framework. Qualitative Report, 21(5), 811–831.

Chan, S. H., Song, Q., Rivera, L. H., & Trongmateerut, P. (2016). Using an educational

computer program to enhance student performance in financial accounting. Journal

of Accounting Education, 36, 43–64. https://doi.org/10.1016/j.jaccedu.2016.05.001

Chang, H. H. (2008). Intelligent agent’s technology characteristics applied to online auctions’

task: A combined model of TTF and TAM. Technovation, 28(9), 564–577.

https://doi.org/10.1016/j.technovation.2008.03.006

Cheung, R., & Vogel, D. (2013). Predicting user acceptance of collaborative technologies:

An extension of the technology acceptance model for e-learning. Computers &

Education, 63, 160–175. https://doi.org/10.1016/j.compedu.2012.12.003

Chung, E., Subramaniam, G., & Christ Dass, L. (2020). Online Learning Readiness Among

University Students in Malaysia Amidst Covid-19. Asian Journal of University

Education, 16(2), 45. https://doi.org/10.24191/ajue.v16i2.10294

117

Collins, K. M. T., Onwuegbuzie, A. J., & Jiao, Q. G. (2007). A Mixed Methods Investigation of

Mixed Methods Sampling Designs in Social and Health Science Research. Journal of

Mixed Methods Research, 1(3), 267–294.

https://doi.org/10.1177/1558689807299526

Conselho de Ministros. (2021a, January 21). Comunicado do Conselho de Ministros de 21

de janeiro de 2021. https://www.portugal.gov.pt/pt/gc22/governo/comunicado-de-

conselho-de-ministros?i=397

Conselho de Ministros. (2021, March 11). Comunicado do Conselho de Ministros de 11 de

março de 2021. https://www.portugal.gov.pt/pt/gc22/governo/comunicado-de-

conselho-de-ministros?i=407

Conselho de Ministros. (2021b, April 15). Comunicado do Conselho de Ministros de 15 de

abril de 2021. https://www.portugal.gov.pt/pt/gc22/governo/comunicado-de-conselho-

de-ministros?i=414

Conselho de Ministros. (2021c). Controlar a Pandemia. https://www.sns.gov.pt/wp-

content/uploads/2021/04/CI_15abr.pdf

Creswell, J. W. (2009). Research design: Qualitative, quantitative, and mixed methods

approaches, 3rd ed (pp. xxix, 260). Sage Publications, Inc.

Croxton, R. A. (2014). The Role of Interactivity in Student Satisfaction and Persistence in

Online Learning. 10(2), 12.

Cucinotta, D., & Vanelli, M. (2020). WHO Declares COVID-19 a Pandemic. Acta Bio-Medica:

Atenei Parmensis, 91(1), 157–160. https://doi.org/10.23750/abm.v91i1.9397

Dakduk, S., González, Á., & Portalanza, A. (2019). Learn About Structural Equation

Modeling in SmartPLS With Data From the Customer Behavior in Electronic

Commerce Study in Ecuador (2017). SAGE Publications, Ltd.

https://doi.org/10.4135/9781526498205

D’Ambra, J., & Wilson, C. S. (2004). Use of the World Wide Web for international travel:

Integrating the construct of uncertainty in information seeking and the task-

118

technology fit (TTF) model. Journal of the American Society for Information Science

and Technology, 55(8), 731–742. https://doi.org/10.1002/asi.20017

D’Ambra, J., Wilson, C. S., & Akter, S. (2013). Application of the task-technology fit model to

structure and evaluate the adoption of E-books by Academics. Journal of the

American Society for Information Science and Technology, 64(1), 48–64.

https://doi.org/10.1002/asi.22757

Davis, F. D. (1989). Perceived Usefulness, Perceived Ease of Use, and User Acceptance of

Information Technology. MIS Quarterly, 13(3), 319–340.

https://doi.org/10.2307/249008

Davis, T., & Frederick, T. V. (2020). The Impact of Multimedia in Course Design on Students’

Performance and Online Learning Experience: A Pilot Study of an Introductory

Educational Computing Course. Online Learning, 24(3).

https://doi.org/10.24059/olj.v24i3.2069

Dennis, A. R., George, J. F., Jessup, L. M., Nunamaker, J. F., & Vogel, D. R. (1988).

Information Technology to Support Electronic Meetings. MIS Quarterly, 12(4), 591–

624. https://doi.org/10.2307/249135

Dishaw, M. T., Eierman, M. A., Iversen, J. H., & Philip, G. (2013). An Examination of the

Characteristics Impacting Collaborative Tool Efficacy: The Uncanny Valley of

Collaborative Tools. Journal of Information Technology Education: Research, 12.

Dishaw, M. T., & Strong, D. M. (1999). Extending the technology acceptance model with

task–technology fit constructs. Information & Management, 36(1), 9–21.

https://doi.org/10.1016/S0378-7206(98)00101-3

Dong, E., Du, H., & Gardner, L. (2020). An interactive web-based dashboard to track

COVID-19 in real time. The Lancet Infectious Diseases, 20(5), 533–534.

https://doi.org/10.1016/S1473-3099(20)30120-1

eportugal.gov.pt. (2021d, January 14). Anunciadas medidas para novo confinamento geral—

EPortugal.gov.pt. eportugal.gov.pt. https://eportugal.gov.pt/noticias/anunciadas-

medidas-para-novo-confinamento-geral

119

Fahey, R. (2020, April 2). 3.9 billion people are locked down due to coronavirus. Mail Online.

https://www.dailymail.co.uk/news/article-8181001/3-9-billion-people-currently-called-

stay-homes-coronavirus.html

Faul, F., Erdfelder, E., Buchner, A., & Lang, A.-G. (2009). Statistical power analyses using

G*Power 3.1: Tests for correlation and regression analyses. Behavior Research

Methods, 41(4), 1149–1160. https://doi.org/10.3758/BRM.41.4.1149

Faul, F., Erdfelder, E., Buchner, A., & Lang, A.-G. (2020). G*Power (3.1.9.7) [Computer

software]. https://www.psychologie.hhu.de/arbeitsgruppen/allgemeine-psychologie-

und-arbeitspsychologie/gpower

Flores, M. A., Simão, A. M. V., Barros, A., Flores, P., Pereira, D., Fernandes, E. L., Ferreira,

P. C., & Costa, L. (2021). Ensino e aprendizagem à distância em tempos de COVID-

19: Um estudo com alunos do Ensino Superior. Revista Portuguesa de Pedagogia,

55, e055001–e055001. https://doi.org/10.14195/1647-8614_55_1

Francescucci, A., & Foster, M. (2013). The VIRI (Virtual, Interactive, Real-Time, Instructor-

Led) Classroom: The Impact of Blended Synchronous Online Courses on Student

Performance, Engagement, and Satisfaction. Canadian Journal of Higher Education,

43(3), 78–91.

Francescucci, A., & Rohani, L. (2019). Exclusively Synchronous Online (VIRI) Learning: The

Impact on Student Performance and Engagement Outcomes. Journal of Marketing

Education, 41(1), 60–69. https://doi.org/10.1177/0273475318818864

Gerhart, N., Peak, D. A., & Prybutok, V. R. (2015). Searching for New Answers: The

Application of Task-Technology Fit to E-Textbook Usage. Decision Sciences Journal

of Innovative Education, 13(1), 91–111. https://doi.org/10.1111/dsji.12056

Goodhue, D., Lewis, W., & Thompson, R. (2006). PLS, Small Sample Size, and Statistical

Power in MIS Research. Proceedings of the 39th Annual Hawaii International

Conference on System Sciences (HICSS’06), 8, 202b–202b.

https://doi.org/10.1109/HICSS.2006.381

120

Goodhue, D., & Thompson, R. (1995). Task-Technology Fit and Individual Performance. MIS

Q. https://doi.org/10.2307/249689

Groccia, J. E. (2018). What Is Student Engagement? New Directions for Teaching and

Learning, 2018(154), 11–20. https://doi.org/10.1002/tl.20287

Gurcan, F., & Cagiltay, N. E. (2020). Research trends on distance learning: A text mining-

based literature review from 2008 to 2018. Interactive Learning Environments, 0(0),

1–22. https://doi.org/10.1080/10494820.2020.1815795

Hair, J. F., Hult, G. T. M., & Ringle, C. M. (2014). A Primer on Partial Least Squares

Structural Equation Modeling (PLS-SEM) (1st ed.). SAGE Publications.

Hair, J. F., Hult, G. T. M., Ringle, C. M., & Sarstedt, M. (2016). A Primer on Partial Least

Squares Structural Equation Modeling (PLS-SEM) (2nd ed.). SAGE Publications.

Hair, J. F., Hult, G. T. M., Ringle, C. M., & Sarstedt, M. (2021). A Primer on Partial Least

Squares Structural Equation Modeling (PLS-SEM) (3rd ed.). SAGE Publications.

Hair, J. F., Ringle, C. M., & Sarstedt, M. (2011). PLS-SEM: Indeed a Silver Bullet. Journal of

Marketing Theory and Practice, 19(2), 139–152. https://doi.org/10.2753/MTP1069-

6679190202

Hair, J. F., Ringle, C. M., & Sarstedt, M. (2013). Partial Least Squares Structural Equation

Modeling: Rigorous Applications, Better Results and Higher Acceptance. Long

Range Planning, 46(1), 1–12. https://doi.org/10.1016/j.lrp.2013.01.001

Hair, J. F., Risher, J. J., Sarstedt, M., & Ringle, C. M. (2019). When to use and how to report

the results of PLS-SEM. European Business Review, 31(1), 2–24.

https://doi.org/10.1108/EBR-11-2018-0203

Harrati, N., Bouchrika, I., & Mahfouf, Z. (2017). Investigating the uptake of educational

systems by academics using the technology to performance chain model. Library Hi

Tech, 35(4), 629–648. https://doi.org/10.1108/LHT-01-2017-0029

Hayes, A. F. (2009). Beyond Baron and Kenny: Statistical Mediation Analysis in the New

Millennium. Communication Monographs, 76(4), 408–420.

https://doi.org/10.1080/03637750903310360

121

Henseler, J., Ringle, C. M., & Sarstedt, M. (2015). A new criterion for assessing discriminant

validity in variance-based structural equation modeling. Journal of the Academy of

Marketing Science, 43(1), 115–135. https://doi.org/10.1007/s11747-014-0403-8

Hidayanto, A. N., & Setyady, S. T. (2014). Impact of Collaborative Tools Utilization on Group

Performance in University Students. Turkish Online Journal of Educational

Technology - TOJET, 13(2), 88–98.

Hodges, C. B., Moore, S., Lockee, B. B., Trust, T., & Bond, M. A. (2020). The Difference

Between Emergency Remote Teaching and Online Learning.

https://vtechworks.lib.vt.edu/handle/10919/104648

Hoehle, H., & Huff, S. (2012). Advancing Task-Technology Fit Theory: A formative

measurement approach to determining task-channel fit for electronic banking

channels. Information Systems Foundations : Theory Building in Information

Systems, 133–169.

Holley, D., & Oliver, M. (2010). Student engagement and blended learning: Portraits of risk.

Computers & Education, 54(3), 693–700.

https://doi.org/10.1016/j.compedu.2009.08.035

Hou, C.-K. (2012). Examining the effect of user satisfaction on system usage and individual

performance with business intelligence systems: An empirical study of Taiwan’s

electronics industry. International Journal of Information Management, 32(6), 560–

573. https://doi.org/10.1016/j.ijinfomgt.2012.03.001

Huda, N. (2011). Distance Education for Rural Agricultural Workers in Indonesia. Asian

Journal of Distance Education, 9(1), 35–45.

INE. (2019). Sociedade da Informação e do Conhecimento—Inquérito à Utilização de

Tecnologias da Informação e da Comunicação pelas Famílias (Sociedade Da

Informação e Do Conhecimento). Instituto Nacional de Estatística.

https://www.ine.pt/xportal/xmain?xpid=INE&xpgid=ine_destaques&DESTAQUESdest

_boui=354447559&DESTAQUESmodo=2&xlang=pt

122

INE. (2020). Sociedade da Informação e do Conhecimento—Inquérito à Utilização de

Tecnologias da Informação e da Comunicação nas Famílias.

https://www.ine.pt/xportal/xmain?xpid=INE&xpgid=ine_destaques&DESTAQUESdest

_boui=415621509&DESTAQUESmodo=2

INE. (2021). Sociedade da Informação e do Conhecimento—Inquérito à Utilização de

Tecnologias da Informação e da Comunicação nas Famílias. Instituto Nacional de

Estatística.

https://www.ine.pt/xportal/xmain?xpid=INE&xpgid=ine_destaques&DESTAQUESdest

_boui=473557834&DESTAQUESmodo=2

Isaac, O., Aldholay, A., Abdullah, Z., & Ramayah, T. (2019). Online learning usage within

Yemeni higher education: The role of compatibility and task-technology fit as

mediating variables in the IS success model. Computers & Education, 136, 113–129.

https://doi.org/10.1016/j.compedu.2019.02.012

Jardina, J. R., Chaparro, B. S., & Abdinnour, S. (2021). Extending the Task-Technology Fit

(TTF) Model to E-Textbook Usage by Students and Instructors. International Journal

of Information and Communication Technology Education (IJICTE), 17(1), 120–137.

https://doi.org/10.4018/IJICTE.2021010108

Jarvis, C. B., MacKenzie, S. B., & Podsakoff, P. M. (2003). A Critical Review of Construct

Indicators and Measurement Model Misspecification in Marketing and Consumer

Research. Journal of Consumer Research, 30(2), 199–218.

https://doi.org/10.1086/376806

Jeong, H., Hmelo-Silver, C. E., & Jo, K. (2019). Ten years of Computer-Supported

Collaborative Learning: A meta-analysis of CSCL in STEM education during 2005–

2014. Educational Research Review, 28, 100284.

https://doi.org/10.1016/j.edurev.2019.100284

Joosten, T., & Cusatis, R. (2020). Online Learning Readiness. American Journal of Distance

Education, 34(3), 180–193. https://doi.org/10.1080/08923647.2020.1726167

123

Kalin, J. (2012). Doing What Comes Naturally? Student Perceptions and Use of

Collaborative Technologies. International Journal for the Scholarship of Teaching and

Learning, 6(1). https://eric.ed.gov/?id=EJ1145195

Khechine, H., Raymond, B., & Augier, M. (2020). The adoption of a social learning system:

Intrinsic value in the UTAUT model. British Journal of Educational Technology, 51(6),

2306–2325. https://doi.org/10.1111/bjet.12905

Kock, N., & Hadaya, P. (2018). Minimum sample size estimation in PLS-SEM: The inverse

square root and gamma-exponential methods. Information Systems Journal, 28(1),

227–261. https://doi.org/10.1111/isj.12131

Koranteng, F. N., Sarsah, F. K., Kuada, E., & Gyamfi, S. A. (2020). An empirical

investigation into the perceived effectiveness of collaborative software for students’

projects. Education and Information Technologies, 25(2), 1085–1108.

https://doi.org/10.1007/s10639-019-10011-7

Laakso, M.-J., Kaila, E., & Rajala, T. (2018). ViLLE--Collaborative Education Tool: Designing

and Utilizing an Exercise-Based Learning Environment. Education and Information

Technologies, 23(4), 1655–1676. https://doi.org/10.1007/s10639-017-9659-1

Laal, M., & Laal, M. (2012). Collaborative learning: What is it? Procedia - Social and

Behavioral Sciences, 31, 491–495. https://doi.org/10.1016/j.sbspro.2011.12.092

Latheef, Z. I., Robinson, R., & Smith, S. (2021). Realistic Job Preview as an Alternative Tool

to Improve Student Readiness for Online Learning. Online Learning, 25(2).

https://doi.org/10.24059/olj.v25i2.2216

Lau, S. L., & Sim, T. Y. (2020). Feedback of University Students on Online Delivery Learning

During the COVID-19 Pandemic Period. 2020 IEEE Conference on E-Learning, e-

Management and e-Services (IC3e), 13–18.

https://doi.org/10.1109/IC3e50159.2020.9288409

Lee, M. K. O., Cheung, C. M. K., & Chen, Z. (2005). Acceptance of Internet-based learning

medium: The role of extrinsic and intrinsic motivation. Information & Management,

42(8), 1095–1104. https://doi.org/10.1016/j.im.2003.10.007

124

Lee, Y. (2006). An empirical investigation into factors influencing the adoption of an e‐

learning system. Online Information Review, 30(5), 517–541.

https://doi.org/10.1108/14684520610706406

Lewin, C., Lai, K.-W., van Bergen, H., Charania, A., Ntebutse, J. G., Quinn, B., Sherman, R.,

& Smith, D. (2018). Integrating Academic and Everyday Learning Through

Technology: Issues and Challenges for Researchers, Policy Makers and

Practitioners. Technology, Knowledge and Learning, 23(3), 391–407.

https://doi.org/10.1007/s10758-018-9381-0

Liaw, S.-S. (2008). Investigating students’ perceived satisfaction, behavioral intention, and

effectiveness of e-learning: A case study of the Blackboard system. Computers &

Education, 51(2), 864–873. https://doi.org/10.1016/j.compedu.2007.09.005

Liaw, S.-S., & Huang, H.-M. (2013). Perceived satisfaction, perceived usefulness and

interactive learning environments as predictors to self-regulation in e-learning

environments. Computers & Education, 60(1), 14–24.

https://doi.org/10.1016/j.compedu.2012.07.015

Lin, H.-C., Han, X., Lyu, T., Ho, W.-H., Xu, Y., Hsieh, T.-C., Zhu, L., & Zhang, L. (2020).

Task-technology fit analysis of social media use for marketing in the tourism and

hospitality industry: A systematic literature review. International Journal of

Contemporary Hospitality Management, 32(8), 2677–2715.

https://doi.org/10.1108/IJCHM-12-2019-1031

Lin, W.-S. (2012). Perceived fit and satisfaction on web learning performance: IS

continuance intention and task-technology fit perspectives. International Journal of

Human-Computer Studies, 70(7), 498–507.

https://doi.org/10.1016/j.ijhcs.2012.01.006

Lu, H.-P., & Yang, Y.-W. (2014). Toward an understanding of the behavioral intention to use

a social networking site: An extension of task-technology fit to social-technology fit.

Computers in Human Behavior, 34, 323–332.

https://doi.org/10.1016/j.chb.2013.10.020

125

Lundvoll Nilsen, L. (2011). Collaborative Work by Using Videoconferencing: Opportunities for

Learning in Daily Medical Practice. Qualitative Health Research, 21(8), 1147–1158.

https://doi.org/10.1177/1049732311405683

M. Klopping, I., & Mckinney, E. (2004). Extending the Technology Acceptance Model and the

Task-Technology Task Technology Fit Model to Consumer E-Commerce. Information

Technology, Learning, and Performance Journal, 22(1).

https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.121.3397&rep=rep1&type

=pdf

Marion, T. J., & Fixson, S. K. (2021). The Transformation of the Innovation Process: How

Digital Tools are Changing Work, Collaboration, and Organizations in New Product

Development*. Journal of Product Innovation Management, 38(1), 192–215.

https://doi.org/10.1111/jpim.12547

Masadeh, R., Tarhini, A., Mohammed, A. B., & Maqableh, M. (2016). Modeling Factors

Affecting Student’s Usage Behaviour of E-Learning Systems in Lebanon.

International Journal of Business and Management, 11(2), p299.

https://doi.org/10.5539/ijbm.v11n2p299

Massey, A. P. (2008). Collaborative Technologies. In F. Burstein & C. W. Holsapple (Eds.),

Handbook on Decision Support Systems 1: Basic Themes (pp. 341–354). Springer.

https://doi.org/10.1007/978-3-540-48713-5_17

McGill, T. J., & Hobbs, V. J. (2008). How students and instructors using a virtual learning

environment perceive the fit between technology and task. Journal of Computer

Assisted Learning, 24(3), 191–202. https://doi.org/10.1111/j.1365-2729.2007.00253.x

McGill, T. J., & Klobas, J. E. (2009). A task–technology fit view of learning management

system impact. Computers & Education, 52(2), 496–508.

https://doi.org/10.1016/j.compedu.2008.10.002

Meroño‐Cerdan, A. L., Soto‐Acosta, P., & López‐Nicolás, C. (2008). Analyzing collaborative

technologies’ effect on performance through intranet use orientations. Journal of

126

Enterprise Information Management, 21(1), 39–51.

https://doi.org/10.1108/17410390810842246

Michaelides, R., Morton, S. C., & Liu, W. (2013). A framework for evaluating the benefits of

collaborative technologies in engineering innovation networks. Production Planning &

Control, 24(2–3), 246–264. https://doi.org/10.1080/09537287.2011.647880

Moore, G. C., & Benbasat, I. (1991). Development of an Instrument to Measure the

Perceptions of Adopting an Information Technology Innovation. Information Systems

Research, 2(3), 192–222. https://doi.org/10.1287/isre.2.3.192

Moser, K. M., Wei, T., & Brenner, D. (2021). Remote teaching during COVID-19:

Implications from a national survey of language educators. System, 97, 102431.

https://doi.org/10.1016/j.system.2020.102431

Murphy, D., & Yum, J. C. K. (1998). Understanding Hong Kong distance learners. Distance

Education, 19(1), 64–80. https://doi.org/10.1080/0158791980190106

Northey, G., Bucic, T., Chylinski, M., & Govind, R. (2015). Increasing Student Engagement

Using Asynchronous Learning: Journal of Marketing Education.

https://doi.org/10.1177/0273475315589814

Novick, G. (2008). Is there a bias against telephone interviews in qualitative research?

Research in Nursing & Health, 31(4), 391–398. https://doi.org/10.1002/nur.20259

Oliveira, T., Faria, M., Thomas, M. A., & Popovič, A. (2014). Extending the understanding of

mobile banking adoption: When UTAUT meets TTF and ITM. International Journal of

Information Management, 34(5), 689–703.

https://doi.org/10.1016/j.ijinfomgt.2014.06.004

Orehovački, T., & Babić, S. (2014). Predicting Students’ Continuance Intention Related to

the Use of Collaborative Web 2.0 Applications. International Conference on

Information Systems Development (ISD).

https://aisel.aisnet.org/isd2014/proceedings/Education/7

Orme, E., Rossiter, N., Rowe, M., & Thomas, L. (2020). Facilitating group work in large

cohorts with collaborative technologies. 26(2), 5.

127

Pal, D., & Patra, S. (2020). University Students’ Perception of Video-Based Learning in

Times of COVID-19: A TAM/TTF Perspective. International Journal of Human–

Computer Interaction, 0(0), 1–19. https://doi.org/10.1080/10447318.2020.1848164

Parsons, J., & Taylor, L. G. (2011). Improving Student Engagement. Current Issues in

Education, 14. https://www.semanticscholar.org/paper/Improving-Student-

Engagement-Parsons-Taylor/61879ed8f54ae9fcf80272e98395f680d0ce8843

Pavlou, P. A., Dimoka, A., & Housel, T. J. (2008). Effective Use of Collaborative IT Tools:

Nature, Antecedents, and Consequences. Proceedings of the 41st Annual Hawaii

International Conference on System Sciences (HICSS 2008), 40–40.

https://doi.org/10.1109/HICSS.2008.136

Pentland, B. (1989). USE AND PRODUCTIVITY IN PERSONAL COMPUTING: AN

EMPIRICAL TEST. ICIS 1989 Proceedings. https://aisel.aisnet.org/icis1989/1

Peters, J. (2020, April 28). Google’s Meet teleconferencing service now adding about 3

million users per day. The Verge.

https://www.theverge.com/2020/4/28/21240434/google-meet-three-million-users-per-

day-pichai-earnings

Picciano, A. (2002). Beyond Student Perceptions: Issues of Interaction, Presence, and

Performance in an Online Course. JALN Volume, 6.

https://doi.org/10.24059/olj.v6i1.1870

Presidência do Conselho de Ministros. (2020, maio). Resolução do Conselho de Ministros

40-A/2020, 2020-05-29. Diário da República Eletrónico. https://dre.pt/pesquisa/-

/search/134889278/details/maximized

Presidente da República Portuguesa. (2020, March 18). Decreto do Presidente da

República 14-A/2020, 2020-03-18: Diário da República n.o 55/2020, 3o Suplemento,

Série I de 2020-03-18. https://dre.pt/home/-/dre/130399862/details/maximized

Qaddumi, B., Ayaad, O., Al-Ma’aitah, M. A., Akhu-Zaheya, L., & Alloubani, A. (2021). The

factors affecting team effectiveness in hospitals: The mediating role of using

128

electronic collaborative tools. Journal of Interprofessional Education & Practice, 24,

100449. https://doi.org/10.1016/j.xjep.2021.100449

Rai, A. (2017). Editor’s comments: Avoiding type III errors: formulating IS research problems

that matter. MIS Quarterly. https://www.semanticscholar.org/paper/Editor%27s-

comments%3A-avoiding-type-III-errors%3A-IS-

Rai/cb94f0fe4455505e4546164aeeaeda7c4015f5f3

Rajab, K. D. (2018). The Effectiveness and Potential of E-Learning in War Zones: An

Empirical Comparison of Face-to-Face and Online Education in Saudi Arabia. IEEE

Access, 6, 6783–6794. https://doi.org/10.1109/ACCESS.2018.2800164

Raven, A., Leeds, E., & Park, C. (2010). Digital Video Presentation and Student

Performance: A Task Technology Fit Perspective. International Journal of Information

and Communication Technology Education, 6(1), 17–29.

https://doi.org/10.4018/jicte.2010091102

Reitoria da ULisboa. (2020, March 9). Comunicados—Reitoria da ULisboa | Faculdade de

Ciências da Universidade de Lisboa: À comunidade da ULisboa (09/03/2020).

https://ciencias.ulisboa.pt/pt/comunicados-reitoria-da-ulisboa

Resta, P., & Laferrière, T. (2007). Technology in Support of Collaborative Learning.

Educational Psychology Review, 19(1), 65–83. https://doi.org/10.1007/s10648-007-

9042-7

Robinson, C. C., & Hullinger, H. (2008). New Benchmarks in Higher Education: Student

Engagement in Online Learning. Journal of Education for Business, 84(2), 101–109.

https://doi.org/10.3200/JOEB.84.2.101-109

Rogers, E. M. (1983). Diffusion of Innovations (SSRN Scholarly Paper ID 1496176). Social

Science Research Network. https://papers.ssrn.com/abstract=1496176

Rojabi, A. R. (2020). Exploring EFL Students’ Perception of Online Learning via Microsoft

Teams: University Level in Indonesia. English Language Teaching Educational

Journal, 3(2), 163. https://doi.org/10.12928/eltej.v3i2.2349

129

Rossman, G. B., & Wilson, B. L. (1985). Numbers and Words: Combining Quantitative and

Qualitative Methods in a Single Large-Scale Evaluation Study. Evaluation Review,

9(5), 627–643. https://doi.org/10.1177/0193841X8500900505

Sandford, A. (2020, April 2). Coronavirus: Half of humanity on lockdown in 90 countries.

Euronews. https://www.euronews.com/2020/04/02/coronavirus-in-europe-spain-s-

death-toll-hits-10-000-after-record-950-new-deaths-in-24-hou

Saragih, A., Adwie, J., & Hendrawan, A. (2021). Determinants and Consequences of

Student Learning Satisfaction During Covid-19 Pandemic. Jurnal Ilmiah Akuntansi

Dan Bisnis, 16, 1–19. https://doi.org/10.24843/JIAB.2020.v16.i01.p01

Sarstedt, M., Ringle, C. M., & Hair, J. F. (2020). Partial Least Squares Structural Equation

Modeling. In C. Homburg, M. Klarmann, & A. E. Vomberg (Eds.), Handbook of

Market Research (pp. 1–47). Springer International Publishing.

https://doi.org/10.1007/978-3-319-05542-8_15-2

Sarstedt, M., Ringle, C. M., Smith, D., Reams, R., & Hair, J. F. (2014). Partial least squares

structural equation modeling (PLS-SEM): A useful tool for family business

researchers. Journal of Family Business Strategy, 5(1), 105–115.

https://doi.org/10.1016/j.jfbs.2014.01.002

Serhan, D. (2020). Transitioning from Face-to-Face to Remote Learning: Students’ Attitudes

and Perceptions of using Zoom during COVID-19 Pandemic. International Journal of

Technology in Education and Science, 4(4), 335–342.

https://doi.org/10.46328/ijtes.v4i4.148

Shakya, S., Sharma, G., & Thapa, K. B. (2017). State Education System with e-learning in

Nepal: Impact and Challenges. Journal of the Institute of Engineering, 13(1), 10–19.

https://doi.org/10.3126/jie.v13i1.20344

Shih, Y.-Y., & Chen, C.-Y. (2013). The study of behavioral intention for mobile commerce:

Via integrated model of TAM and TTF. Quality & Quantity, 47(2), 1009–1020.

https://doi.org/10.1007/s11135-011-9579-x

130

Shmueli, G., Sarstedt, M., Hair, J. F., Cheah, J.-H., Ting, H., Vaithilingam, S., & Ringle, C.

M. (2019). Predictive model assessment in PLS-SEM: Guidelines for using

PLSpredict. European Journal of Marketing, 53(11), 2322–2347.

https://doi.org/10.1108/EJM-02-2019-0189

Singh, V., & Thurman, A. (2019). How Many Ways Can We Define Online Learning? A

Systematic Literature Review of Definitions of Online Learning (1988-2018).

American Journal of Distance Education, 33(4), 289–306.

https://doi.org/10.1080/08923647.2019.1663082

Snyder, H. (2019). Literature review as a research methodology: An overview and

guidelines. Journal of Business Research, 104, 333–339.

https://doi.org/10.1016/j.jbusres.2019.07.039

Soper, D. S. (2022). Post-hoc Statistical Power Calculator for Multiple Regression.

https://www.danielsoper.com/statcalc/calculator.aspx?id=9

Soria, K. M., Chirikov, I., & Jones-White, D. (2020). The Obstacles to Remote Learning for

Undergraduate, Graduate, and Professional Students.

https://escholarship.org/uc/item/5624p4d7

Staples, D. S., & Seddon, P. (2004). Testing the Technology-to-Performance Chain Model.

Journal of Organizational and End User Computing (JOEUC), 16(4), 17–36.

https://doi.org/10.4018/joeuc.2004100102

Steele, C. (2019, February 22). What is the Digital Divide? Digital Divide Council.

http://www.digitaldividecouncil.com/what-is-the-digital-divide/

Sun, J. C.-Y., & Rueda, R. (2012). Situational interest, computer self-efficacy and self-

regulation: Their impact on student engagement in distance education. British

Journal of Educational Technology, 43(2), 191–204. https://doi.org/10.1111/j.1467-

8535.2010.01157.x

Tam, C., & Oliveira, T. (2016a). Performance impact of mobile banking: Using the task-

technology fit (TTF) approach. International Journal of Bank Marketing, 34(4), 434–

457. https://doi.org/10.1108/IJBM-11-2014-0169

131

Tam, C., & Oliveira, T. (2016b). Understanding the impact of m-banking on individual

performance: DeLone & McLean and TTF perspective. Computers in Human

Behavior, 61, 233–244. https://doi.org/10.1016/j.chb.2016.03.016

Taylor, C. W., & Hunsinger, D. S. (2011). A Study of Student Use of Cloud Computing

Applications. 22(3), 15.

Taylor, S., & Todd, P. A. (1995). Understanding Information Technology Usage: A Test of

Competing Models. Information Systems Research, 6(2), 144–176.

Teddlie, C., & Tashakkori, A. (2006). A general typology of research designs featuring mixed

methods. Undefined. https://www.semanticscholar.org/paper/A-general-typology-of-

research-designs-featuring-Teddlie-

Tashakkori/984acce3d901d695d99e7269f2a880ac5874ac0d

Thompson, R. L., Higgins, C. A., & Howell, J. M. (1991). Utilization of Persona/Computers

Personal Computing: Toward a Conceptual Model of Utilization1. MIS Quarterly,

125–143.

Transkriptor. (2021). Transkriptor. Transkriptor. https://transkriptor.com/

Triandis, H. C. (1979). Values, attitudes, and interpersonal behavior. Nebraska Symposium

on Motivation, 27, 195–259.

ULisboa. (2020, March 10). Colaboração a distância [Text]. ULisboa: Serviços de

tecnologias da informação. https://www.ulisboa.pt/info/colaboracao-distancia

UNESCO. (2020a, March 4). 290 million students out of school due to COVID-19: UNESCO

releases first global numbers and mobilizes response. UNESCO.

https://en.unesco.org/news/290-million-students-out-school-due-covid-19-unesco-

releases-first-global-numbers-and-mobilizes

UNESCO. (2020b, March 4). Education: From disruption to recovery. UNESCO.

https://en.unesco.org/covid19/educationresponse

Universidade NOVA de Lisboa. (2020a, March 11). Atualizações COVID-19: Universidade

NOVA de Lisboa suspende aulas presenciais. Universidade NOVA de Lisboa.

https://www.unl.pt/nova/atualizacoes-covid-19

132

Universidade NOVA de Lisboa. (2020b, March 23). Recursos para aulas online (PT/EN).

NOVA FCSH. https://www.fcsh.unl.pt/recursos-para-aulas-online/

Universidade NOVA de Lisboa. (2021, January 14). Decreto n.o 3-A/2021: Diário da

República n.o 9/2021, 1o Suplemento, Série I de 2021-01-14, páginas 5—29.

https://dre.pt/dre/detalhe/decreto/3-a-2021-153959843

Usoro, A., Shoyelu, S., & Kuofie, M. (2010). Task-Technology Fit and Technology

Acceptance Models Applicability to eTourism. Journal of Economic Developmet,

Management, IT, Finance, and Marketing, 2(1), 1–32.

Van De Ven, A. H. (2007). Engaged Scholarship: A Guide For Organizational and Social

Research. Oxford University Press.

Vargo, D., Zhu, L., Benwell, B., & Yan, Z. (2021). Digital technology use during COVID-19

pandemic: A rapid review. Human Behavior and Emerging Technologies, 3(1), 13–

24. https://doi.org/10.1002/hbe2.242

Venkatesh, V., Brown, S. A., & Bala, H. (2013). Bridging the Qualitative-Quantitative Divide:

Guidelines for Conducting Mixed Methods Research in Information Systems. MIS

Quarterly, 37(1), 21–54.

Venkatesh, V., Brown, S., & Sullivan, Y. (2016). Guidelines for Conducting Mixed-methods

Research: An Extension and Illustration. Journal of the Association for Information

Systems, 17(7). https://doi.org/10.17705/1jais.00433

Venkatesh, V., & Davis, F. D. (2000). A Theoretical Extension of the Technology Acceptance

Model: Four Longitudinal Field Studies. Management Science, 46(2), 186–204.

Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User Acceptance of

Information Technology: Toward a Unified View. MIS Quarterly, 27(3), 425–478.

https://doi.org/10.2307/30036540

Venkatesh, V., Thong, J. Y. L., & Xu, X. (2012). Consumer Acceptance and Use of

Information Technology: Extending the Unified Theory of Acceptance and Use of

Technology. MIS Quarterly, 36(1), 157–178. https://doi.org/10.2307/41410412

133

VERBI Software. (2019). MAXQDA 2020 (20.4.1) [Computer software]. Berlin, Germany:

VERBI Software. maxqda.com

Virdyananto, A. L., Dewi, M. A. A., Hidayanto, A. N., & Hanief, S. (2016). User acceptance of

human resource information system: An integration model of Unified Theory of

Acceptance and Use of Technology (UTAUT), Task Technology Fit (TTF), and

Symbolic Adoption. 2016 International Conference on Information Technology

Systems and Innovation (ICITSI), 1–6. https://doi.org/10.1109/ICITSI.2016.7858227

Wan, L., Xie, S., & Shu, A. (2020). Toward an Understanding of University Students’

Continued Intention to Use MOOCs: When UTAUT Model Meets TTF Model. SAGE

Open, 10(3), 2158244020941858. https://doi.org/10.1177/2158244020941858

Watts, L. (2016). Synchronous and Asynchronous Communication in Distance Learning: A

Review of the Literature. Quarterly Review of Distance Education, 17(1), 23–32.

Webster, J., & Hackley, P. (1997). Teaching Effectiveness in Technology-Mediated Distance

Learning. The Academy of Management Journal, 40(6), 1282–1309.

https://doi.org/10.2307/257034

Wei, H.-C., & Chou, C. (2020). Online learning performance and satisfaction: Do perceptions

and readiness matter? Distance Education, 41(1), 48–69.

https://doi.org/10.1080/01587919.2020.1724768

WHO. (2021). A timeline of WHO’s response to COVID-19 in the WHO European Region: A

living document (update to version 2.0 from 31 December 2019 to 31 July 2021).

WHO European Region. https://www.euro.who.int/en/health-topics/health-

emergencies/coronavirus-covid-19/publications-and-technical-guidance/2021/a-

timeline-of-whos-response-to-covid-19-in-the-who-european-region-a-living-

document-update-to-version-2.0-from-31-december-2019-to-31-july-2021

Widiantoro, A. D., & Harnadi, B. (2019). Voluntariness Difference in Adoption of E-Learning

Technology among University Students. 2019 23rd International Computer Science

and Engineering Conference (ICSEC), 402–408.

https://doi.org/10.1109/ICSEC47112.2019.8974819

134

Windows Central. (2020, April 29). Microsoft Teams hits 75 million daily active users.

Windows Central. https://www.windowscentral.com/microsoft-teams-hits-75-million-

daily-active-users

Wu, B., & Chen, X. (2017). Continuance intention to use MOOCs: Integrating the technology

acceptance model (TAM) and task technology fit (TTF) model. Computers in Human

Behavior, 67, 221–232. https://doi.org/10.1016/j.chb.2016.10.028

XXII Governo, R. P. (2020, March 13). Comunicação enviada às escolas sobre suspensão

das atividades com alunos nas escolas de 16 de março a 13 de abril.

https://www.portugal.gov.pt/pt/gc22/comunicacao/documento?i=comunicacao-

enviada-as-escolas-sobre-suspensao-das-atividades-com-alunos-nas-escolas-de-

16-de-marco-a-13-de-abril

Yadegaridehkordi, E., Iahad, N. A., & Ahmad, N. (2014). Task-technology fit and user

adoption of cloud-based collaborative learning technologies. 2014 International

Conference on Computer and Information Sciences (ICCOINS), 1–6.

https://doi.org/10.1109/ICCOINS.2014.6868439

Yadegaridehkordi, E., Shuib, L., Nilashi, M., & Asadi, S. (2019). Decision to Adopt Online

Collaborative Learning Tools in Higher Education: A Case of Top Malaysian

Universities. Education and Information Technologies, 24(1), 79–102.

https://doi.org/10.1007/s10639-018-9761-z

Yi, Y. J., You, S., & Bae, B. J. (2016). The influence of smartphones on academic

performance: The development of the technology-to-performance chain model.

Library Hi Tech, 34(3), 480–499. https://doi.org/10.1108/LHT-04-2016-0038

Younas, M., Chao, L., Sohaib, K., & Abu, B. (2021). Effect Of Home Environment On

Students’ Academic Achievements At Higher Level. Ilkogretim Online - Elementary

Education Online, Vol. 19(3), 3931–3947.

https://doi.org/10.17051/ilkonline.2020.03.735550

Yu, T.-K., & Yu, T.-Y. (2010). Modelling the factors that affect individuals’ utilisation of online

learning systems: An empirical study combining the task technology fit model with the

135

theory of planned behaviour. British Journal of Educational Technology, 41(6), 1003–

1017. https://doi.org/10.1111/j.1467-8535.2010.01054.x

Zainol, Z., Yahaya, N., Mohamat, N. A., & Zain, N. N. B. (2017). Factors Influencing Mobile

Learning Among Higher Education Students in Malaysia. 2(8), 6.

Zawacki-Richter, O., & Naidu, S. (2016). Mapping research trends from 35 years of

publications in Distance Education. Distance Education, 37(3), 245–269.

https://doi.org/10.1080/01587919.2016.1185079

Zeng, X., & Wang, T. (2021). College Student Satisfaction with Online Learning during

COVID-19: A review and implications. International Journal of Multidisciplinary

Perspectives in Higher Education, 6(1), 182–195.

Zhao, X., Lynch, J. G., Chen, Q., & article., J. D. served as editor and G. F. served as

associate editor for this. (2010). Reconsidering Baron and Kenny: Myths and Truths

about Mediation Analysis. Journal of Consumer Research, 37(2), 197–206.

https://doi.org/10.1086/651257

Zhao, Y., Lei, J., Yan, B., Lai, C., & Tan, H. S. (2005). What Makes the Difference? A

Practical Analysis of Research on the Effectiveness of Distance Education. Teachers

College Record, 107(8), 1836–1884. https://doi.org/10.1111/j.1467-

9620.2005.00544.x

Zhou, T., Lu, Y., & Wang, B. (2010). Integrating TTF and UTAUT to explain mobile banking

user adoption. Computers in Human Behavior, 26(4), 760–767.

https://doi.org/10.1016/j.chb.2010.01.013

Zoom. (2020, April 22). 90-Day Security Plan Progress Report: April 22. Zoom Blog.

https://blog.zoom.us/90-day-security-plan-progress-report-april-22/

Zoom. (2021). Zoom for Higher Education. Zoom Video Communications.

https://explore.zoom.us/docs/doc/Zoom%20for%20Higher%20Education.pdf?_ga=2.

250941190.1808753461.1635604045-1989931961.1635604045