Abstract
Background: A systematic review of the literature uncovered little explanation as to why academics as individuals switch mobile applications or continue to use the same mobile applications. Knowing and being able to explain the personal behaviour are important and could be of value to practitioners and researchers, especially when it comes to predicting the future behaviour.
Objectives: The main objective of this study was to explain and to model switching and continued use of mobile computing applications by South African academics.
Method: A survey was carried out within South African academics as participants to test 14 hypotheses and explain switching and continued use of mobile computing applications. The adopted positivist study was associated with the quantitative research approach to test hypotheses so that the truth about objectives are obtainable.
Results: The results of the study contributed to the body of knowledge by explaining and modelling the determinants influencing and significant to mobile computing application switching and continued use behaviour.
Conclusion: The developed model may help to inform practitioners and researchers on what actually satisfies users to switch or continue to use mobile computing applications.
Keywords: academics; continued use; mobile computing applications; social exchange; switching; TTF.
Introduction
Technology is rapidly developing, with current trends suggesting a dramatic rise in the use of mobile computing applications that connect through the Internet (Bergman 2015). Yet, despite their significance and effects, the behaviour of academics towards the use of mobile computing applications and services is not well defined from a theoretical and practical viewpoint. This deters a thorough understanding and explanations for switching and continued use of mobile computing applications or services, especially amongst South African academics.
There are a number of mobile computing applications and services, which include mobile operating systems (Android, iOS, Blackberry, etc.), social media applications (such as WhatsApp, Facebook, Instagram, WeChat, Twitter, Tiktok, LinkedIn, etc.), collaboration applications (such as MS Teams, Zoom, Google meet, Jitsi meet, etc.) and mobile storage application (such as Dropbox, OneDrive, Google drive, Dubox, etc.), which academics may switch around or continue using. The potential life cycle of these mobile applications may be determined by its continued usage. Those who design and sell mobile applications have to keep end-user’s satisfactions higher otherwise, end-users may switch to other mobile computing services or switch back to alternative user-friendly services.
This study took a view that switching behaviour is a significant construct for post acceptance of old and newly developed mobile computing applications and services. The determinants affecting consumer continuance intension is an influential factor of actual usage, thus suggesting that the two phenomena could be studied collectively. Although individuals, especially academics, switch and continue to use mobile computing applications or services, there is a literature gap to explain why they switch or continue to use the same mobile application.
The review of literature has revealed inadequate or lack of explanation towards the switching and continued use behaviour of South African academics. Knowing and being able to explain the behaviour is deemed to be important and could be of value to practitioners and researchers, especially when it comes to predicting the future behaviour. Furthermore, investigations and developing satisfactory mobile computing services may encourage its continued usage. This study aimed to construct the knowledge gap by modelling the determinants that explain the switching and continued use behaviour of academics with respect to mobile computing applications and services.
Background of the research problem
It has become increasingly important to understand the academics’ use of mobile computing applications and services. Many positive effects have resulted from the use of technology, such as social inclusion, improved access to information, day-to-day assistance and innovations for health care (Piwek et al. 2016). However, adverse side effects such as technological addiction, alleged data violations, decreased physical exercise, cyberbullying and poor work-life balance remain common (Ho & Tan 2020). Academics may be avoiding switching into mobile computing because of such negative side effects of unsatisfactory work-life balance from technological addictions.
Theoretical knowledge gaps
Theis study was motivated by the lack of explanations for switching and continued use behaviour of individuals regarding mobile computing applications and services. The review of literature has revealed inadequate models that may inform the switching and continued use behaviour with respect to mobile computing applications amongst professional individuals such as academics.
The theoretical knowledge gap is in the inadequacy or the lack of frameworks or models that explain the switching and continued use behaviour of academics in using mobile computing applications and services.
Practical knowledge gaps
Mobile computing applications’ continued use and switching behaviour are not well explained for practitioners and developers of mobile computing applications. Application developers need to understand switching and continued usage behaviour of individuals to predict future switching or continued use.
The expected life cycle of existing mobile computing applications can be predicted through continued usage or switching. For instance, companies that are developing and selling mobile computing applications must retain high-customer loyalty through positive continued usage.
Literature and study hypothesis
Van der Merwe (2015) developed and tested a theoretical switching intention model with data collected on switching intentions. Switching intention behaviour was investigated that included the relationship of alternative attractiveness, perceived worth and the switching costs. The actual switching behaviour data were then comparable with the theoretical switching intentions, and the results were discussed. Both contexts of switching are examined to determine the role of interpersonal traits. Van der Merwe used online questionnaires for obtaining primary data that were collected from a cross-sectional forum. The participants included people who had or have mobile device contracts with mobile network operator companies. By using Analyses of Moment Structure (AMOS), feature estimates were obtained using maximum likelihood, and confidence intervals were calculated by bootstrapping.
Nimako and Nyame (2015) reviewed a theoretical model in consumer switching behaviour (CSB), and their study reveals that at least 10 theoretical models of CSB already exist. The gap on the existing CSB model is that they do not address issues related to computing research area. Both models focused on the business and marketing research fields, understanding the determinants and mechanism of customer switching behaviour. Although numerous studies have been conducted to enhance understanding of the CSB concept, empirical work in the field of customer switching behaviour is void.
Osah and Kyobe (2015) investigated the factors that influence user retention in M-pesa, a mobile money transfer application. The goal of this study was to (1) develop a research model grounded in theory, (2) identify and discuss factors from the literature that are most likely to influence user continuance intention towards M-pesa and (3) validate the model within the different sampling context to identify the antecedents and determinants of user’s continuance intention towards M-pesa, in Kenya. A literature review, expert pre-testing, pilot testing and statistical validation were used to develop the survey instrument employed to measure the study’s constructs. The findings show that factors such as object, control, attitudinal and behavioural beliefs all influence user persistence intention. The unexpected finding of the rankings of predictive strength of the factors ushers in a new chapter and opens up new avenues of investigation in future studies (Osah & Kyobe 2015).
Shiau and Luo (2013), conducted the study to better understand the factors that influence the intention to continue using a popular hedonic information system, blogs. The expectation–confirmation theory (ECT) has been modified to account for perceived enjoyment, habit and user involvement. An online survey was used to collect data. There were 430 valid responses in total. Structural equation modelling was used to evaluate the structural equation model (SEM). The findings indicate that user involvement, satisfaction and perceived enjoyment all predicted long-term blog use intention. However, there was no significant relationship between habit and satisfaction or use intention.
Based on the given literature, the following study model was conceptualised from by triangulating the three theoretical model that was used as lenses for this study. The three theoretical model underpinned by this study included Information Systems (IS) continuance model, switching intention model and task technology fit (TTF) model. The study hypotheses are shown in Figure 1 and Table 1.
| TABLE 1: Hypotheses, constructs and measuring items. |
The following are the study hypotheses.
The hypotheses and constructs are measured and tested through the measuring items.
Methodology
To explain switching and continued use of mobile computing applications by academics, positivism was identified as the most appropriate paradigm for the study. Blaikie (2007) pointed out that positivist studies are associated with the quantitative research approach that test hypotheses so that the truth about objectives is obtainable.
Questionnaire development and distributions
Questionnaires were used to collect quantitative data from academics. Babbie (2005) posited that a questionnaire is a method explicitly developed to produce knowledge valuable for research. Questionnaires can include close-ended or open-ended questions, and both provide structure to the process of data gathering. The close-ended questions are more specific and less prone to interpretation and verbosity than the open-ended questions (Bryman & Bell 2011). This study used a close-ended questionnaire with the Likert scale method of one to five points. The questionnaire was designed using the SurveyMonkey, and the link was distributed via email addresses, WhatsApp and Instagram. The study also posted a link on LinkedIn application because that is where most professionals, including academics, can be found.
Study participants
This study used a non-probability sample in the form of chain-referral sampling, as the study wanted to use a specific group of participants, academics; however, it was not possible to specify the population. The only known information was that the participants should be working in an academic environment at any South African University. The research is based on their personal choice not as university sanctioned applications. The study sent questionnaires to known academics from different South African Universities and requested those participants to extend the invitation to other academics they know.
Table 2 displays the demographic data from the 220 eventual academic participants. The first section of the survey instrument requested the participants to accept or decline participation in the survey. A total of 216 academics accepted and participated in the survey whilst a total of four participants did not click accept but nevertheless participated. The four participants who participated without clicking the accept button may have not seen the check box because it is on the first page that contained user consent information.
| TABLE 2: Demographic information (gender and age). |
With regard to gender, there were 114 males who participated, which is representing (51.8%) and 106 were females, (48.2%). The age group with most participants was 36–45 years represented by 79 (35.9%) participants, followed by age 25–35 years (22.7%), participants of age below 25 years were 39 (17.7%) and the fourth place was age category 46–55 years, with 33 (15%) participants being the older academic who are above 55 years were 19 (8.6%).
Demographic information
Table 2 shows participants’ demographics results and Table 3 displays their educational qualification and academic specialisation.
| TABLE 3: Participants’ demographic (educational qualification and academic specialisation). |
Table 2 displays the demographic data from the 220 eventual academic participants. The first section of the survey instrument requested the participants to accept or decline participation in the survey. A total of 216 academics accepted and participated in the survey, whilst a total of four participants did not click accept but nevertheless participated. The four participants who participated without clicking the accept button may have not seen the check box because it is on the first page that contained user consent information.
With regard to gender, there were 114 males (51.8%) and 106 females (48.2%), who participated. Most participants were of the age group 36–45 years represented by 79 (35.9%) participants, followed by age 25–35 years (22.7%), participants of age below 25 years were 39 (17.7%) and the fourth place was age category 46–55 years, with 33 (15%) participants then the older academic who are above 55 years were 19 (8.6%).
The majority of participants (76) had master’s degree as their highest qualification and this represented 34.5% of the sampled academics, this was followed by Diploma (45), which represented (22.5%), this sample may be academics who are working as tutors at universities because the study involved tutors and mentors (Table 3). Thirty-two participants hold Bachelor’s degree (14.5%), those with Doctoral degrees were 28 (12.7%) participants, whilst 21 (9.5%) participants selected ‘others’ on educational qualification, which the study assumes could be student mentors and student assistants who are still completing certain academic qualifications. The participant with honours degrees were 18, which constitutes 8.2% of the sample.
The frequency of academic specialisation indicates that majority of the participants (63) were from the Information and communication technology (28.6%), followed by 47 participants from Humanities (21.4%), 27 participants from Education (12.3%) whilst 18 specialised in Commerce and management (18.2%). There are also 16 participants who specialised in Management sciences (7.3%), 14 Engineering (6.4%) and 12 Social sciences (5.5%). Nine participants were from Agriculture and natural resources (4.1%), six from Health and medical sciences (2.7%), five from Law (2.3%), the least was two from Fine and performing arts (0.9%). One participant selected ‘others’, which constitutes only 0.5% of the sample.
Study results and analysis
This section describes and explains the analyses of the data 0.439), H12 (p = 0.405) and H14 (p = 0.464). Some hypotheses obtained through the research instrument. The statistical were dropped because their p-value is > 0.05, indicating that it analysis of the data were performed using SPSS 26.0 (Pallant 2020) and was not significant enough to be supported as shown in Table 4.
| TABLE 4: The decision made for each hypothesis. |
The final model for switching and continued use
The research model showed 14 hypotheses to explain academic’s behaviour towards switching and continued use of mobile computing applications. Table 4 displays the hypotheses formulated based on TTF, IS continuance and switching intention model together with an outcome for each hypothesis, on whether they were supported.
According to Ogee et al. (2015) and Mcleod (2019), if the p-value is ≤ 0.05, it is significant because it indicates strong evidence concerning hypothesis, it indicates that there is 95% confidence level concerning the results. If a p-value is > 0.05, there is no significant relationship between the constructs, which means the hypothesis is rejected.
As displayed in Table 4, there were hypotheses that were not supported and were dropped in order to develop the final research model. However, the five hypotheses supported are H1 (p-value of 0.013), H3 (p-value of 0.000), H9 (p-value of 0.000), H10 (p-value of 0.000) and H13 (p-value of 0.000) because their p-value is less than or equal to 0.05, indicating that the value was significant enough to be supported. The following hypotheses were not supported: H2 (p-value of 0.243), H4 (p-value of 0.526), H5 (p-value of 0.922), H6 (p-value of 0.013), H7 (p-value of 0.622), H8 (p-value of 0.638), H11 (p-value of 0.439), H12 (p = 0.405) and H14 (p = 0.464). Some hypotheses were dropped because their p-value is > 0.05, indicating that it was not significant enough to be supported as shown in Table 4.
The final contracts of SWITCON behaviour model shows only the five factors that are significant in influencing the switching and continued use of mobile computing applications amongst South African academics:
H1: Technological characteristics may have positive effects towards MCA switching behaviour.
H3: Task characteristics may have positive effects towards MCA switching behaviour.
H9: Perceived usefulness may have positive effects towards MCA switching behaviour.
H10: Perceived usefulness may have positive effects towards continued use of MCA.
H13: Social norm may have positive effects towards MCA switching behaviour.
Discussion and recommendations
This study confirmed that the technological characteristics of the mobile computing application have a positive effect on the switching behaviour of South African academics. The hypothesis was supported. As predicted through the research model, technological characteristics had a positive influence on the MCA switching behaviour. According to Chen et al. (2012), users’ switching behaviour is influenced by the technology characteristics of the application.
 |
FIGURE 2: Recommended model for switching and continued use of mobile computing applications. |
|
The result of this study concurs with Nel and Boshoff (2013), who found that specialists using video conferencing to operate a patient need quite a lot of technological support to achieve their goal if they are not technologically oriented, similarly with academics. The result of this study reveals that technological characteristics have a significantly positive effect on MCA switching behaviour amongst South African academics.
The results of this study, however, did not confirm if the technological characteristics have positive effects on the continued use of MCA amongst the South African academics. This hypothesis was not supported. According to the research model, technological characteristics may have positive effects towards continued use of MCA in the South African academics. However, technological characteristics towards continued use of MCA in South Africa do not have a positive influence on academics. This result is, however, inconsistent with TTF theory.
The results confirmed that the task characteristics have a positive effect on the MCA switching behaviour. This hypothesis was supported significantly. The results confirm the prediction of the research model, which is consistent with the proposition of TTF model. The results of hypothesis H3 are consistent with Goodhue (1995) who showed that task characteristics are important in determining the appropriateness of technology. Likewise, academic tasks and characteristics do influence the switching behaviour of South African academics.
The result did not confirm that the task characteristics have a positive effect on the continued use of MCA. This hypothesis was not supported, which did not confirm the prediction of the research model.
The results of hypothesis H4 are not consistent with Goodhue (1995) who claimed that task characteristics are important towards determining the appropriateness of technology. Likewise, academic tasks and characteristics do not influence the decision to continue using a computing application, amongst South African academics.
The study did not confirm that an individual’s characteristics have a positive effect on the MCA switching behaviour. This hypothesis was not supported. This result is not consistent with the TTF theory. According to Nan (2011), users’ adoption and use of a technology is also not determined by an individual’s characteristics. The individual’s skills and efficacy (individual characteristics) to utilise mobile computing applications is not significant, and thus does not affect the academic’s switching behaviour.
The results of this study did not confirm that the individual characteristics have positive effect on the continued use of MCA amongst South African academics. This hypothesis was not supported. This result is not consistent with the TTF model, which posited that an individual’s characteristics are important to determine the appropriate use of a technology.
The study did not confirm that satisfaction characteristics have a positive effect on the MCA switching behaviour amongst South African academics. This hypothesis was not supported. As predicted in the research model, satisfaction characteristics do not have a positive influence on the MCA switching behaviour. According to (Osah & Kyobe 2015), the use of technology by users is determined by satisfaction characteristics. Thus, an academic’s expectations, needs for using new mobile computing applications or the pleasure derived from using existing mobile computing applications significantly affect the mobile application switching behaviour.
As predicted in the research model, satisfaction characteristics do not have a positive influence in the continued use behaviour amongst South African academics. This result is not consistent with the Information System Continuance model. Users’ continued use of technology is determined by satisfaction characteristics (Osah & Kyobe 2015). This means that academics are likely to not continue to use mobile computing applications if they are satisfied with its offering.
The study confirmed that the perceived usefulness characteristics have a positive effect on the MCA switching behaviour amongst South African academics. This hypothesis was supported. That is, as predicted in the research model, perceived usefulness characteristics have a positive influence MCA switching behaviour. This result is consistent with what the switching intention model posits. According to Lim et al. (2019), users’ switching behaviour is determined by perceived usefulness characteristics. It is, thus, inferred that trust, application quality and service quality is what drives the academics’ usefulness perception for switching to new mobile computing applications.
The results of this study did confirm that perceived usefulness characteristics have a positive effect on the continued use of MCA amongst South African academics. This hypothesis was supported.
In line with Lim et al. (2019), users’ continued use of a technology is determined by the perceived usefulness characteristics. The usefulness perception of the existing mobile computing applications is also a significant aspect that affects the continued use behaviour of academics.
The study did not confirm that the switching costs characteristics have a positive effect on the MCA switching behaviour amongst South African academics. This hypothesis was not supported. This result is not consistent with IS continued use and switching intention models. According to Ray et al. (2012), the characteristics of users’ technology switching behaviour is often determined by cost characteristics. This means that academics may not weigh the cost involved with using an application relative to the new application they may want to switch to.
The study confirmed that cost characteristics have a positive effect on the MCA continued use behaviour amongst South African academics. This hypothesis was supported. This result is again consistent with IS continued use and switching intention models. Continued use behaviour is also determined by costs associated with the use. This means that academics may weigh the cost involved with using an application relative to the new application they may want to use now.
The study confirmed that the social norm characteristics have a positive effect on the MCA switching behaviour amongst South African academics. The hypothesis was supported as predicted in the research model. According to Ajzen and Fishbein (1975), the adoption of technology by users may be determined by social norms. This means that social norms may actually influence academics to switch to a new mobile computing application
The study did not confirm that the social norm characteristics have a positive effect on the MCA continued use behaviour amongst South African academics. The hypothesis was not supported as predicted in the research model. This result is again not consistent with Ajzen and Fishbein (1975), who posited that the use of technology may be determined by social norms. This means that social norms actually do not influence academics to continue using a mobile computing application.
Study contribution
This section outlines the contributions made from this study. They are discussed as a response to the theoretical and practical knowledge gaps.
Theoretical contributions
The contribution of this study towards theory is that it merged the appropriate information systems literature in order to enhance what is known regarding the use of mobile computing applications. This means that the study brought out the factors that explain why individuals switch between mobile applications and why they may continue using mobile computing applications. By triangulating three theoretical frameworks, the study showed how the theories of TTF, IS continuance and switching intention model could best explain the behaviour of individuals towards switching and continued use of mobile computing applications. The model drawn from the study provides a theoretical knowledge that bridges the gaps identified from the review of the existing literature.
Practical contribution
The study provides a model for switching and continued use of mobile computing applications. The model may be used by mobile application developers, retailers and service providers in predicting the future use of these applications. Amongst other things, this may help practitioners to identify why and which applications are appropriate for the tasks, what to do to prevent or encourage switching to a new application, and what factors are significant for individuals to continue using the mobile computing applications they already have.
Conclusion
The study explains the switching and continued use of mobile computing applications amongst academics as individuals. Although there has been a proliferation of the use of mobile computing applications and services, the research in what explains the switching and continued use behaviour amongst individuals is still at infancy. Switching behaviour is an important factor for post adoption of newly developed mobile computing applications, and the intention to continue using the application is an influential factor towards actual usage. This then suggested that the two phenomena should be studied concurrently to have a better behavioural explanation. Although individual academics use different mobile computing applications, there is little explanation as to why they switch mobile applications or continue to use the same mobile application. This study bridges these knowledge gaps.
The review of literature has revealed an inadequate or a lack of explanation to the switching and continued use behaviour of academics as individuals, regarding mobile computing applications. Knowing and being able to explain this behaviour was deemed to be important and could be of value to both practitioners and researchers, especially when it comes to predicting the future behaviour. The study has identified the factors that explain the switching and continued use behaviour, thus bridging the theoretical and practical knowledge gaps.
Acknowledgements
This article is partially based on the A.T. Kgopa’s thesis for the degree of Doctor of Computing in Informatics at the Tshwane University of South Africa, South Africa, with supervisor Prof. R.M. Kekwaletswe and Dr A. Pretorius.
Competing interests
This research study was not funded and no publication prohibitions, conditions or limitations were placed on the researcher.
Authors’ contributions
R.M.K. contributed as a supervisor and A.P. as a co-supervisor. Both provided comments on the work and redirected me to the study. They played an important role in data collection and helped to make sure that A.T.K. followed the right procedure to complete my study.
Ethical considerations
This article followed all ethical standards for a research without direct contact with human or animal subjects.
Funding information
The funding received was for language editing, printing, fees and was funded for A.T. Kgopa by Tshwane University of Technology Capacity Development Grands.
Data availability
The authors’ data collection accession code is CJ3RHKY, and the following is the link for proof of data collected using SurveyMonkey: https://www.surveymonkey.com/r/CJ3RHKY.
Disclaimer
The views and opinions expressed in this article are those of the authors and do not necessarily reflect the official policy or position of any affiliated agency of the authors.
References
Ajzen, I. & Fishbein, M., 1975, ‘A Bayesian analysis of attribution processes’, Psychological Bulletin 82(2), 261–277. https://doi.org/10.1037/h0076477
Babbie, E.R., 2005, The basics of social research, 3rd edn., Wadsworth Publishing Company, Belmont, CA.
Bergman, N., 2015, Surviving the techstorm: Strategy in times of technological uncertainty, Lid Publishing.
Blaikie, N., 2007, Approaches to social inquiry, 2nd edn., Polity, Cambridge.
Bryman, A. & Bell, E., 2011, ‘Ethics in business research’, Business Research Methods 7(5), 23–56.
Chen, L., Meservy, T. & Gillenson, M., 2012, ‘Understanding information systems continuance for information-oriented mobile applications’, Communications of the Association for Information Systems 30(9), 127–146. https://doi.org/10.17705/1CAIS.03009
Goodhue, D.L., 1995, ‘Understanding user evaluations of information systems’, Management Science 41(12), 1827–1844. https://doi.org/10.1287/mnsc.41.12.1827
Ho, K.W. & Tan, M.K.T., 2020, ‘Challenges to social mobility in Singapore’, The Singapore Economy: Dynamism and Inclusion, 221–280
Lim, S.H., Kim, D.J., Hur, Y. & Park, K., 2019, ‘An empirical study of the impacts of perceived security and knowledge on continuous intention to use mobile fintech payment services’, International Journal of Human–Computer Interaction 35(10), 886–898. https://doi.org/10.1080/10447318.2018.1507132
Nan, N., 2011, ‘Capturing bottom-up information technology use processes: A complex adaptive systems model’, MIS Quarterly 35(2), 505–532. https://doi.org/10.2307/23044054
Nel, J. & Boshoff, C., 2013, ‘Cell phone banking adoption and the continuance of use in an internet-banking context: A study of consumer cognitive evaluations’, PhD, Dept. of Business Management, Marketing Management: University of Pretoria.
Nimako, S.G. & Nyame, R.K., 2015, ‘Influence of demography, religiosity and porting behaviour on switching behaviour of mobile subscribers: Evidence from a developing country’, International Journal of Business and Management 10(7), 158–171. https://doi.org/10.5539/ijbm.v10n7p158
Ogee, A. & Ellis, M., 2015, What can yo u say when your P-Value is greater than 0.05, The Minitab Blog, Pennsylvania.
Osah, O. & Kyobe, M., 2015, ‘User continuance intention towards mobile technology-enabled services in developing regions’, PhD, Dept. of Information System, University of Cape Town.
Pallant, J., 2020, SPSS survival manual: A step by step guide to data analysis using IBM SPSS, Routledge.
Piwek, L., Ellis, D.A., Andrews, S. & Joinson, A., 2016, ‘The rise of consumer health wearables: Promises and barriers’, PLoS Medicine 13(2), e1001953. https://doi.org/10.1371/journal.pmed.1001953
Ray, S., Kim, S.S. & Morris, J.G., 2012, ‘Research note-online users’ switching costs: Their nature and formation’, Information Systems Research 23(1), 197–213. https://doi.org/10.1287/isre.1100.0340
Shiau, W.L. & Luo, M.M., 2013, ‘Continuance intention of blog users: The impact of perceived enjoyment, habit, user involvement and blogging time’, Behaviour & Information Technology 32(6), 570–583. https://doi.org/10.1080/0144929X.2012.671851
Van der Merwe, M.C., 2015, ‘A comparison between switching intension and switching behavior in the South African mobile telecommunications industry’, DCom Marketing Management, University of Pretoria.
Annexure A
Research Questionnaire
Section B
| 
Hide
Show all
Email this article (Login required)
