The Journal of International Management Studies, Volume 11 Number 1, February, 2016 110 An Examination of Mobile Application Use Intention through the Unified Theory of Acceptance and Use Technology Model Mu-Cheng Wu, Assistant Professor, Physical Education Office, National Chin-Yi University of Technology, Taiwan ABSTRACT This study aims to explore the causal relationships related to use intention for mobile application by investigating users in Northern, Central, and Southern Taiwan with purposive sampling, which selects 400 users as research subjects for a questionnaire survey. In total, 390 copies were collected with a return rate reaching 97.5%; among them, 370 copies were valid, reaching an effective rate of 94.8%. The data acquired was archived with the table coding of SPSS 20.0. AMOS 20.0 was then adopted to analyze the causal relationship between variables. Research results are shown below: 1. performance expectancy did not significantly influence use intention; 2. effort expectancy did not significantly influence use intention; 3. social influence significantly influenced use intention; 4. benefits significantly influenced user behavior; 5. system quality significantly influenced use intention; 6. information quality significantly influenced use intention; 7. service quality significantly influenced use intention; 8. use intention significantly influenced user’s behavior; 9. there was a significant interference effect for gender on seven dimensions of use intention, including performance expectancy, effort expectancy, social influence, benefits, system quality, information quality, and service quality; 10. there was a significant interference effect caused by age in social influence on use intention; and 11. there was a significant interference effect caused by experience in benefits and system quality on use intention. Keywords: Unified Theory of Acceptance and Use of Technology (UTAUT), Information Systems Success Model, mobile application INTRODUCTION With the advancement of information technology in recent years, most relevant research theories on utilization and acceptance of new information technology have focused on factors that raise use intention and acceptance level regarding new information technology and products. Hence, scholars in different research fields have proposed different perspectives and influential factors. Therefore, many variables generated in this manner were able to comprehensively explain the acceptance levels of new information technology. In 1985, Ajzen added perceived behavioral control to the Theory of Reasoned Action (TRA), and proposed the planned behavior theory (TPB), while in 1989, Davis integrated TRA and TPB into a new framework referred to as the Technology Acceptance Model (TAM) with the purpose of discussing the acceptance of new technology among users, describing the acceptance behavior of users towards new information technology, and analyzing the influence of new information technology on the majority of users. Davis thought that “perceived usefulness” and “perceived ease of use” of technology influenced users’ attitude and willingness to use that technology. In his theory, “perceived usefulness” referred to the fact that users subjectively perceive the system as useful to improve work performance. “Perceived ease of use” was the result of higher user confidence towards self-efficacy and control, and a more active and
12
Embed
An Examination of Mobile Application Use Intention … Mu-Cheng Wu.pdfwhile AVE should be higher than 0.5 (Fornell and Larcker 1981) to indicate convergent validity. This study conducted
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
The Journal of International Management Studies, Volume 11 Number 1, February, 2016110
An Examination of Mobile Application Use Intention through the Unified Theory of Acceptance and Use Technology Model
Mu-Cheng Wu, Assistant Professor, Physical Education Office,
National Chin-Yi University of Technology, Taiwan
ABSTRACT
This study aims to explore the causal relationships related to use intention for mobile application
by investigating users in Northern, Central, and Southern Taiwan with purposive sampling, which selects
400 users as research subjects for a questionnaire survey. In total, 390 copies were collected with a
return rate reaching 97.5%; among them, 370 copies were valid, reaching an effective rate of 94.8%. The
data acquired was archived with the table coding of SPSS 20.0. AMOS 20.0 was then adopted to analyze
the causal relationship between variables. Research results are shown below: 1. performance expectancy
did not significantly influence use intention; 2. effort expectancy did not significantly influence use
intention; 3. social influence significantly influenced use intention; 4. benefits significantly influenced
user behavior; 5. system quality significantly influenced use intention; 6. information quality significantly
influenced use intention; 7. service quality significantly influenced use intention; 8. use intention
significantly influenced user’s behavior; 9. there was a significant interference effect for gender on seven
dimensions of use intention, including performance expectancy, effort expectancy, social influence,
benefits, system quality, information quality, and service quality; 10. there was a significant interference
effect caused by age in social influence on use intention; and 11. there was a significant interference
effect caused by experience in benefits and system quality on use intention.
Keywords: Unified Theory of Acceptance and Use of Technology (UTAUT), Information Systems Success
Model, mobile application
INTRODUCTION
With the advancement of information technology in recent years, most relevant research theories on
utilization and acceptance of new information technology have focused on factors that raise use intention
and acceptance level regarding new information technology and products. Hence, scholars in different
research fields have proposed different perspectives and influential factors. Therefore, many variables
generated in this manner were able to comprehensively explain the acceptance levels of new information
technology. In 1985, Ajzen added perceived behavioral control to the Theory of Reasoned Action (TRA),
and proposed the planned behavior theory (TPB), while in 1989, Davis integrated TRA and TPB into a
new framework referred to as the Technology Acceptance Model (TAM) with the purpose of discussing
the acceptance of new technology among users, describing the acceptance behavior of users towards new
information technology, and analyzing the influence of new information technology on the majority of
users. Davis thought that “perceived usefulness” and “perceived ease of use” of technology influenced
users’ attitude and willingness to use that technology. In his theory, “perceived usefulness” referred to the
fact that users subjectively perceive the system as useful to improve work performance. “Perceived ease
of use” was the result of higher user confidence towards self-efficacy and control, and a more active and
The Journal of International Management Studies, Volume 11 Number 1, February, 2016 111
positive attitude towards the technology when it was easy to operate and use. The research field that
examines the technology acceptance behavior of users produced several theoretical models derived from
information systems, psychology, and sociology. This led to difficulty for researchers in choosing an
appropriate research model. Inevitably, researchers were forced to select a few dimensions from these
models to create new ones, or select one of the existing models, for relevant studies. As a result,
Venkatesh et al. (2003) combined various theories, including TRA, TPB, TAM, Social Cognitive Theory
(SCT), Model of Personal Computer Utilization (MPCU), Motivational Model (MM), Innovation
Diffusion Theory (IDT), and the integrated TAM and TPB (i.e. C-TAM-TPV) to develop the Unified
Theory of Acceptance and Use of Technology (UTAUT), which contained four main dimensions. This
theory was proved to have an explanatory power of more than 70% after empirical analysis, more
effective than any previous models.
As demonstrated in the above discussion, UTAUT serves as a behavioral measurement model for
managers prior to introducing a new technology. This model can be modified and evaluated based on the
use of information systems, system interface, factors from the surrounding environment, and professional
organizations and technologies.
Businesses are required to spend huge amounts of money and time introducing information systems;
thus, scholars in the relevant fields of information management have long studied relevant issues. In the
meanwhile, they have continuously sought relevant factors concerning positive influences in
informational system success models. The purpose is to improve the effectiveness that the information
system can bring to users. In 1992, DeLone and McLean reviewed literature relevant to information
systems. After compiling 180 articles from seven key information system journals, they produced an
Information System Success Model designed to positively influence businesses. This model included six
dimensions: System Quality, Information Quality, Use, User Satisfaction, Individual Impact, and
Organizational Impact.
With the advancement of technology, an information system brings influences not only to internal
information personnel, but also to the whole business organization. Pitt et al. (1995) argued that the
Information System Success Model of DeLone and McLean (1992) did not thoughtfully consider the role
of services. They therefore proposed the addition of a “Service Quality” dimension to the two original
dimensions: Information Quality and System Quality, which jointly influence use and user satisfaction in
the information system. In 2003 DeLone& McLean modified the information system success model
proposed in 1992 and, taking the suggestion of Pitt et al. (1995), added “service quality” to create a new
and more comprehensive information system success model that consisted of six dimensions: system
quality, information quality, service quality, use, user satisfaction, and net benefits. Among these, system
quality, information quality, and service quality influence use and user satisfaction, while use and user
satisfaction are interdependent and have an influence on net benefits. When the information system
generates performance and benefits, it also affects use intention and user satisfaction.
To sum up the above literature review, the purpose of the UTAUT model lies in the acceptance of
users towards new information technology, the explanation of acceptance behavior for new information
technology demonstrated by users, and analysis of the influence of new information technology on the
majority of users. Additionally, three dimensions: system quality, information quality, and service quality
in the informational system success model directly influence user satisfaction and willingness. Therefore,
the combination of UTAUT model and informational system success model facilitates a discussion on
acceptance willingness and the use of new information technology perceived by users. This study
examines user willingness and use on the part of mobile application program users, and takes the UTAUT
The Journal of International Management Studies, Volume 11 Number 1, February, 2016112
model as its main structure, as well as the addition of information quality, system quality, and service
quality from the informational system success model. Its research results can serve as a reference for the
development and design of future mobile application programs conducted by the Sports Administration of
the Ministry of Education.
RESEARCH METHOD
Research hypothesis Through an examination of previous literature we find that Venkatesh et al. (2003) integrated TRA,
TPB, TAM, SCT, MPCU, MM, IDT, and C-TAM-TPB to produce the UTAUT that unifies four
dimensions including performance expectancy, effort expectancy, social influence, and facilitating
conditions. Upon empirical analysis, this theory demonstrated an explanatory power of more than 70%.
As a result, this study uses the performance expectancy, effort expectancy, social influence, and
facilitating conditions of UTAUT, as well as the system quality, information quality, and service quality
of the informational system success model to propose its research hypotheses:
Hypothesis 1: Performance expectancy significantly influences use intention, while moderating variables
include gender and age.
Hypothesis 2: Effort expectancy significantly influences use intention, while moderating variables
include gender and age.
Hypothesis 3: Social influence significantly influences use intention, while moderating variables include
gender, age and experience.
Hypothesis 4: Facilitating conditions significantly influence use intention, while moderating variables
include age and experience.
Hypothesis 5: System quality significantly influences use intention, while moderating variables include
gender, age and experience.
Hypothesis 6: Information quality significantly influences use intention, while moderating variables
include gender, age and experience.
Hypothesis 7: Service quality significantly influences use intention, while moderating variables include
gender, age and experience.
Hypothesis 8: Use intention significantly influences user behavior.
Research tool This study adopts UTAUT of Venkateshet al. (2003) as its main analysis structure and introduces
dimensions from the informational system success model including “system quality,” “information
quality,” and “service quality” to the structure for analysis. For questionnaire contents, this study mainly
referred to questions proposed by Cheng (2008) about the use of on-line shopping websites by consumers.
Necessary modifications were made according to research needs. The scale consists of five factor
dimensions: performance expectancy, effort expectancy, social influence, facilitating conditions, and use
intention, with a total of 18 questions. Additionally, the three dimensions of system quality, information
quality, and service quality, with 11 total questions, were also modified according to research needs. This
study used Likert’s seven-point scale to distribute points from 1 to 7 according to answers ranging from
“highly disagree” to “highly agree.”
The Journal of International Management Studies, Volume 11 Number 1, February, 2016 113
Research subject
This study aims to investigate the causal relationship of use intention of app iSports Go among
users in Northern, Central, and Southern Taiwan. Utilizing purposive sampling, 400 users were randomly
selected as the research samples for the questionnaire survey. In total, 390 copies were returned with a
return rate of 97.5%; among them, 370 copies were effective, thus producing an effective rate of 94.8%.
Information processing and analysis
Data acquired from effective copies of questionnaires in this study was archived with the table
coding of SPSS 20.0, and AMOS 20.0 was adopted to analyze the relationship between variables.
RESEARCH RESULT
Offending Estimate This study used seven indices, including χ2(Chi-square), χ2,degree of freedom, GFI, AGFI,
RMSEA, CFI and PCFI to examine goodness of fit. Bagozzi and Yi (1988) suggested that offending
estimate should be conducted before test of goodness of fit. Discrepancy function value is between 0.02
and 0.12 while standardized coefficient is between 0.43 and 0.94 not exceeding the standardized value of
0.95. Thus, this study has no offending estimate (Hair, Anderson, Tatham & Black, 1998), and overall
goodness of fit can be examined.
Analysis of measurement model This study used CFA for questionnaire reliability and validity, and also referred to modification
indices (MI) to delete questions (Chen, 2007). This study deleted questions from Effort 4, Facilitating 1,
Facilitating 2, Intention 4, and Behavior 1 in the UTAUT Scale, as well as System 1, Information 1, and
Information 2 in the Informational Success Model. In the end, the reliability and validity of the study was
re-examined.
(1)Test of convergent validity
Convergent validity can be measured from composition reliability (CR) and average variance
extracted (AVE) for dimension. It is suggested CR value should be higher than 0.7 (Bagozzi and Yi, 1988)
while AVE should be higher than 0.5 (Fornell and Larcker 1981) to indicate convergent validity. This
study conducted convergent validity testing of respective dimensions of the UTAUT model, as well as the
informational success model, including performance expectancy, effort expectancy, social influence,
facilitating conditions, use intention, user behavior, system quality, information quality, and service
quality. Factor loading for all dimensions lies between 0.73-0.94, CR between 0.87-0.93, and AVE
between 0.65-0.87, (as shown in Table 1 and 2) which meets the standards of Bagozzi and Yi (1988),
Fornell & Larcker (1981), and Hair, Anderson, Tatham, & Black (1998) and shows convergent validity
for this study.
The Journal of International Management Studies, Volume 11 Number 1, February, 2016114
Lower Upper Lower Upper System Quality < - > Information Quality 0.81 0.76 0.86 0.76 0.86 System Quality < - > Service Quality 0.74 0.68 0.79 0.68 0.79
Information Quality < - > Service Quality 0.74 0.68 0.80 0.68 0.80 (3)Structural Model Analysis
This study took the suggestions of Bagozzi and Yi (1988), Wu (2009), Chen (2007), Hsu (2010), and
Hair et al. (1998) as reference and used seven indices, including χ2 test, the ratio of χ2 to degree of free
freedom, GFI, AGFI, root mean square error of approximation (RMSEA), comparative fit index (CFI), and
PCFI for the determination of goodness of fit for this study’s overall model. As shown in Table 5, Bagozzi& Yi
(1988) suggested the ratio of χ2 and degree of freedom should be smaller than 3. This study, after modification,
achieves a ratio of 2.32. Hair et al. (1998) also advised that the closer GFI and AGFI values are to 1, the better.
The values for this study were 0.90 and 0.86 respectively after modification. Browne and Cudeck (1993)
pointed out a better value of RMSEA is between 0.05 and 0.08, and the modified value of this study is 0.06.
CFI tolerance standard should be higher than 0.90, and this study modifies the value to 0.96, while the
modified value of PCFI is 0.79 in comparison to tolerance, which is higher than 0.50. The research results of
this study show the overall goodness of fit falls within the range of standard acceptance.
Table 5: Goodness of Fit Analysis of Overall Model Index Tolerance Scope Modification Model Determination of Goodness of Fit χ2(Chi-square) The smaller the better 570.64 Ratio of χ2 to Degree of Freedom <3 2.32 Pass GFI >0.80 0.90 Pass AGFI >0.80 0.86 Pass RMSEA <0.08 0.06 Pass CFI >0.90 0.96 Pass PCFI >0.50 0.79 Pass
The Journal of International Management Studies, Volume 11 Number 1, February, 2016116
Diag 1: UTAUT Model
The Journal of International Management Studies, Volume 11 Number 1, February, 2016 117
Table 6: Confirmatory Result of Research Hypothesis Hypothesis Path Relationship Path Value Does hypothesis stand?
1 Performance Expectancy→ Use Intention -0.093 No 2 Effort Expectancy →Use Intention -0.079 No 3 Social Influence→ Use Intention 0.133 Yes 4 Facilitating Conditions→ User Behavior 0.068 No 5 System Quality→ Use Intention 0.451 Yes 6 Information Quality →Use Intention -0.188 Yes 7 Service Quality→ Use Intention 0.717 Yes 8 Use Intention→ User Behavior 0.763 Yes
*p <0.05
(5) Influence of moderating variable
Based on the Chi-square Test proposed by Bentler and Bonnet (1980), the moderating variable of
this study’s research model was tested. When the degree of freedom for the two models are 1 and α=0.05,
chi-square value reaches 3.84, which indicates the existence of a moderating effect. From Tables 8 and 9,
we discover that the CMIN values for this research with respect to the social influence of ages on use
intention, facilitating conditions on user behavior, and experience of system quality on behavior and
intention are 10.3472, 10.7830, and 8.1945 respectively, and all higher than the standardized value of
3.84. Thus, there is a moderating effect.
Table 7: Confirmatory Result of Moderating Effect of Gender on Research Hypothesis
Model DF CMIN P NFI
Delta-1 IFI
Delta-2 RFI
rho-1 TL1 rho2
Does the hypothesis stand?
Mode1 Number 2 Performance Expectancy →Use Intention
1 2.5842 0.1079 0.0003 0.0003 0.0001 0.0001 No
Mode1 Number 2 Effort Expectancy →Use Intention
1 0.0018 0.9666 0.0000 0.0000 -0.0002 -0.0003 No
Mode1 Number 2 Social Influence →Use Intention
1 0.0317 0.8588 0.0000 0.0000 -0.0002 -0.0003 No
Mode1 Number 2 System Quality →Use Intention
1 1.7838 0.1817 0.0002 0.0002 0.0000 0.0000 No
Mode1 Number 2 Information Quality →Use Intention
1 0.3840 0.5355 0.0000 0.0000 -0.0002 -0.0002 No
Mode1 Number 2 Service Quality →Use Intention
1 0.0807 0.7763 0.0000 0.0000 -0.0002 -0.0003 No
Table 8: Confirmatory Results for Moderating Effect of Age on Research Hypothesis
Model DF CMIN P NFI
Delta-1IFI
Delta-2RFI
rho-1 TL1 rho2
Does the hypothesis stand?
Mode1 Number 2 Performance Expectancy →Use Intention
3 6.3665 0.0951 0.0007 0.0007 0.0002 0.0002 No
Mode1 Number 2 Effort Expectancy →Use Intention
3 0.8853 0.8290 0.0001 0.0001 -0.0005 -0.0006 No
Mode1 Number 2 Social Influence →Use Intention 3 10.3472 0.0158 0.0011 0.0012 0.0007 0.0008 Yes Mode1 Number 2 Facilitating Conditions
→ User Behavior 3 6.4394 0.0921 0.0007 0.0008 0.0002 0.0002 No
Mode1 Number 2 System Quality →Use Intention 3 0.7759 0.8552 0.0001 0.0001 -0.0005 -0.0006 No Mode1 Number 2 Information Quality
→ Use Intention 3 1.4743 0.6882 0.0002 0.0002 -0.0004 -0.0005 No
Mode1 Number 2 Service Quality →Use Intention 3 2.0206 0.5681 0.0002 0.0002 -0.0004 -0.0004 No
The Journal of International Management Studies, Volume 11 Number 1, February, 2016118
Table 9: Confirmatory Results for Moderating Effect of Experience on Research Hypothesis
Model DF CMIN P NFI
Delta-1IFI
Delta-2RFI
rho-1 TL1 rho2
Does the hypothesis stand?
Mode1 Number 2 Performance Expectancy →Use Intention
3 2.8883 0.4092 0.0003 0.0003 -0.0004 -0.0005 No
Mode1 Number 2 Effort Expectancy →Use Intention
3 0.9526 0.8127 0.0001 0.0001 -0.0007 -0.0008 No
Mode1 Number 2 Social Influence →Use Intention
3 10.7830 0.0130 0.0011 0.0012 0.0006 0.0006 Yes
Mode1 Number 2 System Quality →Use Intention
3 8.1945 0.0422 0.0008 0.0009 0.0002 0.0003 Yes
Mode1 Number 2 Information Quality →Use Intention
3 4.8162 0.1858 0.0005 0.0006 -0.0002 -0.0002 No
Mode1 Number 2 Service Quality →Use Intention
3 4.0936 0.2515 0.0004 0.0005 -0.0003 -0.0003 No
RESULTS AND DISCUSSION
This study is based on UTAUT and adds three dimensions from the D&M Informational System
Success Model: system quality, information quality, and service quality to discuss user attitude and
behavioral intention towards the internal variables of popular mobile app iSports Go. The research results
are as follows:
(1) No significant impact of performance expectancy on use intention
As shown in the confirmatory results of this study, performance expectancy has no significant
influence on use intention for popular mobile app, iSportsGo. It is assumed that when promoting this
mobile app, iSports Go, the Sport Administration, MOE would like to provide users with an
easy-to-understand interface and simple practical function design with the aim of enhancing the use
intention of users towards it. However, the influence of performance expectancy on use intervention
might have changed due to the interface’s lack of memory function. This led to the unavailability of
services such as the identification of users’ personal information, automatic detection of heart beats, and
the improvement of interpersonal interaction. This research result corresponds with research findings of
Chen (2012) and Chang (2012).
(2) No significant impact for effort expectancy on use intention
As shown in the confirmatoryresults of this study, effort expectancy has no significant impact on the
popular mobile app, iSportsGo. It is assumed this result might be due to users’ excessive expectations
concerning function, interface, and easiness for the mobile app. Most users believed the mobile app,
iSports Go was supposed to help users improve exercise effectiveness, but there was no practical effect in
this regard. Thus, users of the mobile app, iSports Go did not feel it was effective. This research result
matches research findings of previous scholars such as Chen (2012), Hsu (2011), Huang & Huang (2008),
and Gu (2012).
(3) A significant impact of social influence on behavioral intention
The Confirmatory results of this study find social influences significantly impacted behavioral
intention towards mobile app, iSportsGo. It is assumed that, due to current trend of technology becoming
part of our lives, the majority of people are familiar with, and rely on, the convenience of apps. They use
various types of apps anytime and anywhere. Smart phones are integrated into people’s lives. Therefore,
apps downloaded by users also impact the assessment of users in the eye of their friends and families.
When users find the app, iSports Go, improves their images among friends and families, they demonstrate
The Journal of International Management Studies, Volume 11 Number 1, February, 2016 119
higher use intention towards the app. This research finding echoes those of Chen (2012), Hsu (2011),