Page 1
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
Page 2
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
Page 3
I dedicate this work to my family
Page 4
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.
Page 5
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
Page 6
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
Page 7
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
Page 8
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
Page 9
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
Page 10
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
Page 11
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
Page 12
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
Page 13
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).
Page 14
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.
Page 15
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
Page 16
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
Page 17
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
Page 18
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.
Page 19
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
Page 20
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),
Page 21
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.
Page 22
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
Page 23
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’
Page 24
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
Page 25
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
Page 26
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
Page 27
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
Page 28
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
Page 29
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,
Page 30
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
Page 31
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).
Page 32
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).
Page 33
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.
Page 34
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).
Page 35
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
Page 36
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
Page 37
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.
Page 38
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.
Page 39
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.
Page 40
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
Page 41
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
Page 42
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:
Page 43
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.
Page 44
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).
Page 45
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).
Page 46
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
Page 47
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
Page 48
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.
Page 49
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.
Page 50
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
Page 51
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
Page 52
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
Page 53
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
Page 54
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.
Page 55
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)
Page 56
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).
Page 57
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)
Page 58
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)
Page 59
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.
Page 60
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
Page 61
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
Page 62
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.
Page 63
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
Page 64
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
Page 65
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
Page 66
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.
Page 67
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
Page 68
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
Page 69
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
Page 70
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
Page 71
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
Page 72
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.
Page 73
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.
Page 74
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
Page 75
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
Page 76
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.
Page 77
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
Page 78
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).
Page 79
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,
Page 80
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).
Page 81
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
Page 82
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%
Page 83
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
Page 84
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."
Page 85
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.”
Page 86
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
Page 87
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
Page 88
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
Page 89
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
Page 90
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
Page 91
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
Page 92
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%
Page 93
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
Page 94
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
Page 95
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.
Page 96
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.
Page 97
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
Page 98
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%
Page 99
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.
Page 100
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
Page 101
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.
Page 102
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
Page 103
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.
Page 104
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.
Page 105
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
Page 106
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
Page 107
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%
Page 108
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.
Page 109
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.
Page 110
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
Page 111
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.
Page 112
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).
Page 113
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
Page 114
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.
Page 115
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.
Page 116
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.
Page 117
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
Page 118
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
Page 119
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.
Page 120
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).
Page 121
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,
Page 122
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
Page 123
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.
Page 124
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’
Page 125
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.
Page 126
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
Page 127
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.
Page 128
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
Page 129
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-
Page 130
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
Page 131
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
Page 132
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
Page 133
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
Page 134
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
Page 135
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
Page 136
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
Page 137
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
Page 138
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.
Page 139
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
Page 140
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
Page 141
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
Page 142
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
Page 143
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
Page 144
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
Page 145
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
Page 146
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
Page 147
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