Mobile Payment Continuance Intention

Mobile Payment Continuance Intention Frank Bivar Franque A thesis submitted in partial fulfillment of the requirements for the degree of Doctor in Information Management November 2021 NOVA Information Management School Universidade Nova de Lisboa Information Management Specialization in Information Technologies II Doctoral Programme in Information Management Professor Doutor Tiago André Gonçalves Félix de Oliveira, Co-Supervisor! Professor Doutor Carlos Tam Chuem Vai, Co-Supervisor III Doctoral Programme in Information Management Copyright © by Frank Bivar Franque All rights reserved ! IV Doctoral Programme in Information Management Abstract The disruptive development of information and communication technologies over the last two decades has revolutionized the mobile phone industry, exponentially increased the number of mobile phone users, and encouraged companies to make various services available through a mobile phone. Mobile payment is one of the fastest growing services, enabling users to perform financial transactions over a mobile phone. The exponential growth of mobile payment has affected a number of sectors including finance and technology, thus reinforcing the need for a deep understanding of the impact of the continued use of mobile payment services. With this dissertation we contribute to a better understanding of the determinants of continuance intention to use mobile payment at the individual level. For this reason, were developed four studies, one literature review, and three empirical studies. In the first study (Chapter 2) we conducted a literature review of existing studies on individual continuance intention to use an information system. In Chapter 3 we assessed the continuance intention to use m-payment employing two theoretical models, the DeLone and McLean information system success model (D&M ISSM) and the expectation-confirmation model (ECM) in an African context. The impact of task technology fit (TTF) and overall trust on ECM to explain the continuance use of mpayment is analysed in Chapter 4. In the last study, Chapter 5, we assess the impact of culture on continuance intention to use m-payment, combining the ECM and Hofstede’s cultural dimensions. This dissertation provides several contributions for research and practice, contributing to the advancement of knowledge and implications for service managers, service providers, users, and researchers. The literature review applies meta-analysis and weight analysis from 115 empirical studies from continuance intention to use an information system (IS). The findings reveal that the factors with strongest influence on continuance intention to use an IS are affective commitment, attitude, satisfaction, hedonic value, and flow. Moreover, sample size, individualism, uncertainty avoidance, and long-term orientation moderate the relationship of perceived usefulness on continuance intention. Power distance, masculinity, and indulgence moderate the V Doctoral Programme in Information Management relationship of satisfaction on continuance intention. From the first empirical study we examine the influence individual performance drivers on continuance intention to use m–payment in an African context. We find that the most important predictors of continuance intention to use m-payment are individual performance, use, and satisfaction. The second empirical study integrates TTF and overall trust theories and evaluates their relationships for continuance intention to use mobile payment. Findings show that use, individual performance, overall trust, and the moderation role of satisfaction are the most important constructs to explain continuance intention. The last empirical study assesses the impact of culture on m-payment continuance intention. The findings reveal that the relationships between confirmation on satisfaction and perceived usefulness, and perceived usefulness on continuance intention are moderated by uncertainty avoidance. Keywords: Mobile payment, continuance intention, ECM, D&M ISSM, TTF, trust, culture, African context. VI Doctoral Programme in Information Management Resumo O desenvolvimento disruptivo das tecnologias de informação e comunicação nas últimas duas décadas revolucionou a indústria da telefonia móvel, aumentando exponencialmente o número de utilizadores de telemóveis, encorajando desta forma as empresas a disponibilizar diferentes serviços através de um telemóvel. O serviço pagamento móvel é um dos serviços que se encontra em um rápido crescimento permitindo aos utilizadores efetuar transações financeiras através de um telemóvel. O crescimento exponencial do serviço de pagamento móvel tem afetado diferentes sectores, tais como finanças e tecnologia, reforçando a necessidade de uma compreensão profunda do impacto da utilização contínua dos serviços de pagamento móvel. Com o desenvolvimento desta dissertação, esperamos contribuir para uma melhor compreensão dos determinantes da intenção de continuar a usar o serviço de pagamento móvel a nível individual. De forma a concretizar este objetivo foram desenvolvidos um total de quatro estudos distintos. No primeiro estudo (Capítulo 2) realizámos uma revisão bibliográfica dos estudos existentes sobre a intenção de continuar a utilizar um sistema de informação. No capítulo três, avaliámos a intenção de continuar a utilizar o serviço de pagamento móvel, empregando dois modelos teóricos, o DeLone and McLean information system success model (D&M ISSM) e o expectation-confirmation model (ECM) num contexto africano. O impacto do task technology fit (TTF) e o overall trust no modelo ECM para explicar o uso contínuo do serviço de pagamento móvel foi analisado no capítulo quatro. No último estudo, capítulo cinco, avaliámos o impacto da cultura na intenção de continuação da utilização do serviço de pagamento móvel, combinando as dimensões culturais de Hofstede e o modelo ECM. Esta dissertação apresenta várias contribuições para a investigação e para a prática, contribuindo para o avanço do conhecimento, provocando implicações para gestores de serviços, prestadores de serviços, utilizadores e investigadores. O estudo da revisão bibliográfica aplicou meta-analysis e weight analysis a partir de 115 estudos empíricos de intenção continuar a utilizar um sistema de informação (SI). Os resultados revelam que os fatores com maior influência na intenção de continuação da utilização de um SI foram o compromisso afetivo, atitude, satisfação, valor hedónico, e flow. Além disso, VII Doctoral Programme in Information Management o tamanho da amostra, individualismo, prevenção da incerteza, e orientação a longo prazo moderam a relação entre perceção da utilidade e intenção de continuar, distância do poder, masculinidade e indulgência moderam a relação entre satisfação e intenção de continuar. Para o primeiro estudo empírico, examinámos a influência dos fatores de desempenho individual na intenção de continuação da utilização do m-pagamento num contexto africano. Verificámos que os preditores mais importantes da intenção de continuar a utilizar o serviço de pagamento móvel são o desempenho individual, uso e a satisfação. O segundo estudo empírico integrou as teorias da TTF e da confiança geral e avaliou as suas relações para a intenção de continuação da utilização do pagamento móvel. Os resultados mostram que o uso, desempenho individual, confiança geral, o papel de moderação da satisfação são os fatores relevantes para explicar a intenção de continuar a utilizar o serviço de pagamento móvel. O último estudo empírico avalia o impacto da cultura sobre a intenção de continuação do pagamento móvel. Os resultados revelam que as relações entre confirmação, perceção de utilidade com satisfação, perceção de utilidade com intenção de continuar são moderadas pela prevenção da incerteza. Palavras-chave: Pagamento móvel, intenção de continuar, ECM, D&M ISSM, TTF, Confiança, Cultura, contexto africano. VIII Doctoral Programme in Information Management Publications List of studies resulting from the dissertation. Papers (published): Franque, F.B., Oliveira, T., Tam, C. and Santini, F.d.O. (2021), "A meta-analysis of the quantitative studies in continuance intention to use an information system", Internet Research, 31(1), 123-158. https://doi.org/10.1108/INTR-03-2019-0103 Franque, F. B., Oliveira, T., & Tam, C. (2021). Understanding the factors of mobile payment continuance intention: empirical test in an African context. Heliyon, 7(8). https://doi.org/10.1016/j.heliyon.2021.e07807 Papers (under review or submitted) Franque, F. B., Oliveira, T., & Tam, C. (2021). Continuance intention of mobile payment: TTF model with Trust in an African context. Franque, F. B., Oliveira, T., & Tam, C. (2021). Role of the uncertainty avoidance culture on the expectation confirmation model: Mobile payment case. IX Doctoral Programme in Information Management Acknowledgements To my family, especially my father, Gilberto Franque, and to all my brothers, who encouraged me from the beginning to pursue a PhD, and for their inexhaustible support for this journey, mainly emotional support, bearing in mind that my family is in Mozambique, in Tete city. To Prof. Tiago Oliveira, supervisor of this dissertation, for being always available, for all the help, support, guidance, and ideas that significantly contributed to the success of the journey. A huge thanks for everything. To Prof. Carlos Tam, co-supervisor of this dissertation, for being always available, for all the help, support, guidance, and ideas that significantly contributed to the success of the journey. A huge thanks for everything. To Prof. Fernando de Oliveira Santini for all the help, suggestions, and critiques, which contributed to the enrichment of Chapter 2. To Nova Information Management School, for the opportunity given to participate in the PhD program. To Engineering School, Catholic University of Mozambique, for the opportunity given to go to Lisbon to participate in the PhD program. To the new friends from Lisbon, for the good moments of fun, that helped a lot for the success of the journey. To my girlfriend, Gisela Manhique, who helped me tremendously in the discussions and translation from Portuguese to English of the different studies. A big thanks for everything. To me, for the courage to go to a completely new country, with new challenges, for the tireless effort during the five years of the journey. To all my sincere thanks. Chapter 1 – Introduction 1.1. Motivation Mobile phone networks are booming and almost all are interconnected, making people able to communicate and share information anywhere in the world. Associated with this, mobile phone usage is growing exponentially, motivating companies to deliver services via mobile phones (Karjaluoto et al., 2019; Persaud & Azhar, 2012). Taking into consideration that it can be used anywhere and anytime, adding more value to services, new services are being made available via mobile phones, such as m-banking, m-payment, and m-commerce, thereby bringing customers closer to companies and strengthening their relationship (Oliveira et al., 2016). In our work we study Mobile payment (m-payment). M–payment is a payment method that uses mobile phones to make financial transactions such as paying for goods or services, transferring money, and/or withdrawing money (Fan, Shao, Li, & Huang, 2018; Zhou, 2013). Whereas in some regions financial institutions are far from the population, forcing people to travel long distances to use financial services, m-payment has become one of the prominent services (Gao & Waechter, 2015; Zhou, 2014), enabling users to make transactions anytime and anywhere (Lu, Wei, Yu, & Liu, 2017; Zhou, 2014). Society has undergone a disruptive revolution with the entrance of this technology, which affects payment ecosystems. This service is growing exponentially all over the world and is bringing benefits to users and providers. Considering its benefits, companies realize its potential, and are providing it in different ways all over the world (Fan et al., 2018; Singh, Sinha, & Liébana-cabanillas, 2020). M-payment technology originated in the United States and spread throughout the world (Fan et al., 2018). In Africa, this technology was launched in Kenya and was quickly adopted by other countries. Mozambique is one of the African countries that has adopted this technology, helping rural people who do not have banking infrastructure near their homes (Batista & Vicente, 2018). M-payment has a major impact on developing countries as it launches basic financial services such as money transfer, payment of goods and services, and/or withdrawing money, thereby improving people’s lives (Humbani & Wiese, 2018). 2 Doctoral Programme in Information Management A great deal of research has been developed to understand m-payment in different regions (e.g., Chen & Li, 2017; Oliveira, Thomas, Baptista, & Campos, 2016; Sinha, Majra, Hutchins, & Saxena, 2019). However, there are few studies to understand mpayment in an African context (e.g., Chen & Li, 2017; Lin, Featherman, & Sarker, 2017). According to Nabavi et al. (2016) and Shaikh and Karjaluoto (2015) no studies about information system continuance intention were found in the African context. For Bhattacherjee, (2001) the early stage of information systems (IS) adoption is a vital step toward the success of IS, but permanent usage of IS and its success is associated with continued use instead of first usage. In this sense, understanding what factors influence an individual to continue using m-payment has become necessary and important for researchers and companies (Bhattacherjee, 2001; Shaikh & Karjaluoto, 2015). It is thus extremely important to understand the most important drivers that influence continuous use of mobile payment in an African context. 1.2. Continuance intention models Studies on information system continuance intention (ISCI) have used a wide range of theories in combination with the expectation-confirmation model (ECM) (Bhattacherjee, 2001) such as expectation-confirmation theory (ECT) (Oliver, 1986), technology acceptance model (TAM) (Davis, 1989), unified theory of acceptance and use of technology (UTAUT) (Venkatesh et al., 2003), and Flow theory (Getzels & Csikszentmihalyi, 1978), to name a few. However, continuance intention of IS refers to factors that contribute to IS usage for a long time. It involves understanding the longterm factors that contribute to the success of the IS (A. Bhattacherjee, 2001; K. Wang, 2015). In the context of ISCI, ECM was the first theory proposed by (Bhattacherjee, 2001). ECM proposes that satisfaction to use IS is a crucial factor that impacts continuance intention, followed by perceived usefulness of the IS. Also, confirmation of the expectations and perceived usefulness are important factors that influence user satisfaction. After ECM appeared many researchers tested and joined it with other models in different regions and with distinct technologies (Carillo, Scornavacca, & Za, 2017; Hadji & Degoulet, 2016; Hong, Tai, Hwang, Kuo, & Chen, 2017b; Hsiao, Chang, 3 Doctoral Programme in Information Management & Tang, 2016; Zheng, 2019). Most studies used ECM as a base theory. Some used ECM alone (Alraimi et al., 2015; Susanto et al., 2016), and others integrated it with other theories and self-constructs (Chen et al., 2013; Lee, 2010; Limayem & Cheung, 2008). In our work we integrate ECM with DeLone & McLean information system success (D&M ISSM) (DeLone & Mclean, 2003), Task Technology Fit (Goodhue & Thompson, 1995), overall trust (Oliveira et al., 2017), and culture (Hofstede, 1984). 1.3. Research focus Mobile payment services are today becoming more useful and more present in people’s daily lives, especially now with the COVID-19 pandemic. Conducting transactions through a mobile phone is already a reality in people’s lives. Understanding the main factors that affect the intention to continue using m-payment is the focus of this dissertation. IS related areas such as e-learning, internet banking, and e-commerce are not within the scope of this work. The study addresses only the individual level of continuance intention. M-payment is defined as a payment method in which a mobile phone is used to perform financial value exchanges (initiate, authorize, and confirm) anytime and anywhere in return for goods and services (Kujala et al., 2017; Liébana-cabanillas et al., 2018; Shao et al., 2019). There are different ways to conduct a transaction using m-payment. The simplest way is short-message-based, by which the user can make payments using a simple mobile phone (Singh et al., 2020; T. Zhou, 2013). 1.4. Main objectives To understand the most important factors of the intention to continue using m-payment, we divide our work into different studies, each presented in separate chapters. The first study (Chapter 2) addresses weight and meta-analysis of ISCI. This study synthesizes the results of previous studies on ISCI, identifying the most used significant relationships in the literature, and the most studied regions and technologies, thus contributing to the state-of-the-art. 4 Doctoral Programme in Information Management In Chapter 3 we analyse the impact of individual performance for the purpose of continuing to use m-payment. We integrate D&M ISSM and ECM in an African context. In Chapter 4 we analyse the impact of task and technology characteristics, overall trust, and the role of satisfaction as a moderator on continuance intention to use m-payment, considering that trust is an important factor that can influence the usage of m-payment. As financial transactions are sensitive, it is important to have trust in the system. In Chapter 5 the role of culture in the ECM model is assessed, considering that culture can be an important factor in the adoption and intention to continue using m-payment. Our motivation is to understand the impact of culture as a moderator on the ECM model. In Chapter 6 we present a summary of the studies, their implications, and recommendations for future studies. 1.5. Methods According to the literature, there are three main epistemological perspectives (positivism, interpretivism, and realism). In this dissertation we follow the positivism perspective (Smith, 2006). Several methods were applied in this research because we developed a new and different model based on existing theories to understand continuance intention to use mobile payment. The development of the theories was followed by tests in order to explain the subject. A set of hypotheses were developed and empirically tested. 1.5.1. Theoretical frameworks The ECM (Bhattacherjee, 2001) is used in all the empirical studies, from Chapters 3 to 5. Chapter 3 is based on ECM combined with D&M ISSM. Chapter 4 is based on the integration of ECM, TTF, and overall trust, and Chapter 5 is based on ECM integrated with the cultural moderator (Hofstede, 1984). 5 Doctoral Programme in Information Management 1.5.2. Quantitative research methods In all of the studies we use a cross-section online survey design to analyse the main factors of mobile payment continuance intention. The data collection was conducted in Mozambique. In Chapter 3 we describe the collection of 338 valid responses, in Chapter 4 we examine a sample of 384, and in Chapter 5 apply a mixed-methods approach based on 384 valid responses triangulated with field interviews. The data were collected from July 2018 to January 2019. Structure equation modelling (SEM) was used to empirically test the research models. We used PLS-SEM (partial least squares – structural equation modelling) via Smart PLS 3 software (Ringle et al., 2015). 1.6. Path of research This dissertation reports the collection of different interrelated studies on the intention to continue using m-payment, presented separately in different chapters. Some of the studies are already published in international journals with a double-blinded review process; others are submitted for publication and are in different stages of review and preparation. The stage of each study is presented in the Table 1.1. The major conclusions of the studies made from Chapter 2 to 5 are presented at the end of the dissertation. Table 1.1 - Research studies stages Chapter Study name Journal Current stage 2 A Meta-analysis of the quantitative studies in continuance intention to use an information system Internet Research Published 3 Understanding the factors of continuance intention to use mobile payment: Empirical test of the D&M IS success with ECM in the African context Heliyon Published 4 Continuance intention of mobile payment: TTF model with Trust in an African context Submitted to a journal of quartile one of Scimago index Under review 6 Doctoral Programme in Information Management 5 Role of the uncertainty avoidance culture on the expectation confirmation model: mobile payment case Submitted to a journal of quartile one of Scimago index Submitted 7 Doctoral Programme in Information Management 8 Doctoral Programme in Information Management Chapter 2 – A meta-analysis of the quantitative studies in continuance intention to use information systems 2.1. Introduction The evolution of information systems (IS) has offered the opportunity to different types of institutions to advance the capability, efficiency, and responsibility of their services and products, thereby streamlining the day to day activities of their customers (Laukkanen 2007). Its use is growing exponentially, considering the usefulness of IS in society. However, the adoption of IS is not enough to keep them in the market, continuous use is necessary (Bhattacherjee, 2001), to bring a return on the investments that companies make, and to help the users in their activities. Continuous use of IS refers to the decision of the user to continue to use the IS. This behaviour is noted after the user has the first experience with IS. Understanding what factors influence an individual to continue to use IS has become necessary for researchers and companies (A Bhattacherjee, 2001; Shaikh & Karjaluoto, 2015). In recent years, the number of studies in continuance intention to use an IS has grown rampantly and now covers several subjects such as continuance intention in mobile banking services, mobile payment, e-learning, social networking, health applications, e-government, mobile commerce, among others. Considering that the number of studies is growing, different technologies, theories and contexts are being studied, there is plenty of scattered information and different results. With that much information in the background, the process of searching for studies became more difficult, and the need for comprehensive and synthesised information about IS continuance intention became essential. Therefore, it is crucial and necessary to highlight, summarise and clarify the results of existing studies in order to provide a comprehensive picture of continuing to use IS (Fettke, 2006). This process enables theory development and reveals new relationships and gaps (Hamari & Keronen, 2017). There are some literature reviews on IS continuous intention (e.g., Bhattacherjee and Barfar, 2011; Shaikh and Karjaluoto, 2015; Nabavi et al., 2016), that explore different aspects of prior studies such as theories, technologies, and used contexts. However, most of them are narrative and descriptive; none of them has used meta-analysis. This study will use metaand weight 9 Doctoral Programme in Information Management analysis to derive for more empirical results. The meta-analysis is a process of summarising, evaluating, and analysing quantitative research findings (L. Zhang et al., 2012), even if the outcome is non-significant or inconsistent, it can contributes to a pooled conclusion, reinforcing the general validity of the interpretations (Hamari & Keronen, 2017; K. Wu et al., 2011). According to earlier research meta-analysis and weight analysis are considered appropriate methods to review empirical data (Baptista & Oliveira, 2016; Rana et al., 2015; Schmidt & Hunter, 2016; Y. Zhao et al., 2018). We describe the most critical variables in the field, using findings reported in existing research combined with weight analysis of the constructs to identify the best predictors (Y. Zhao et al., 2018) to highlight the best predictors of continuance intention to use an IS, improvements of theories, and the strength of the variables. According to our knowledge, no research addresses: (i) meta-analysis combined with weight analysis in the context of continuance intention to use an IS, or (ii) temporal analysis to understand the evolution of the theoretical models over time. This study can extrapolate broader theoretical implications relating to the positioning and understanding of IS. Contributing to the research, we illustrate the most used relationships, best predictors, most used technologies during a period, the evolution of the number of papers per year, and the evolution of the theoretical model. The overall variables to be used to predict continuance intention to use an IS were illustrated. Beyond synthesising the main findings of the studies, we also created models to understand the temporal evolution of the constructs better. Additionally, we analysed possible moderators in the relationship between perceived usefulness and satisfaction on continuance intention to use IS. The article is organised as follows: in Section 2, is presented the literature review, Section 3 describes the research methodology; in the next section, we present the results of the research followed by a discussion of the findings; the conclusion and future recommendations follow this section. 16 Doctoral Programme in Information Management 2.3.1. Moderator analysis Concerning moderators, we investigated some moderating effects in different relationships. To select the relationships, we used relationships that have enough observations (> 30) (Geyskens et al., 2009; Lipsey & Wilson, 2001). Regarding the moderator variables, we have selected different moderators suggested by the literature (Hofstede & Minkov, 2010; F. D. O. Santini et al., 2019). For our study, we have used methodological, economic, and cultural moderators. This analysis is essential for the literature because it provides researchers a better understanding of the effects of the relationships and provides an overview of the potential moderators that can influence the continuance intention to use IS in future studies. Appendix D presents the moderators, the descriptions, and the coding structure. Sample size was analysed as a methodological moderator, considering that it plays a significant role in varying the effect sizes in the studies (Fern & Monroe, 1996). While a small sample size is more homogeneous, this aspect tends to overestimate the effect size of the relationships (Rosenthal & Rubin, 1982). We have also analysed moderators in the economic context, (1) economic development, and (2) innovation level. Economic development can play an important role as it tends to promote different levels of use of IS. Therefore, we expect that developed economies influence the behaviour of the intention to continue using IS, compared to developing economies (Y. Kim & Peterson, 2017). On regards innovation level, it is considered a potential moderator because it can influence the relationships, taking into consideration that countries with a high level of innovation tend to continue using IS as they have good skills and familiarity with systems usage (Y. Kim & Peterson, 2017). Finally, we have analysed six cultural moderators from Hofstede, power distance, individualism, masculinity, uncertainty avoidance, long term orientation, and indulgence. These cultural dimensions are recognised as the leading indicators of people's beliefs and values that impact their behaviour, so we consider potential moderators that can influence the intention to continue using IS (Hofstede & Minkov, 2010). Our analysis was supported on a hierarchical linear meta-analysis. This analysis uses the multivariate regression format for the variables included in the model and is widely used in meta-analytic research (Geyskens et al., 2009; F. D. O. Santini et al., 2019). 17 Doctoral Programme in Information Management 2.4. Findings 2.4.1. Meta-Analysis Table 2.1 represents the meta-analysis and weight analysis of the 60 relationships that were most often used, and which have occurred three or more times across the 115 studies. Columns 4 to 8 (meta-analysis information) of Table 2.1 present the number of times that a relationship was analysed (total), the sum of samples (sample), an average of the correlation coefficient (AVG of cc), normal standard deviation (Z – value), and 95% confidence interval. In addition, we then show the relationship between the dependent constructs and the independent constructs. Several dependent constructs relate to different independent constructs, such as continuance intention, which relates to 16 different independent constructs, followed by satisfaction, which applies to 14 different constructs and perceived usefulness, which applies to 7 different constructs. The meta-analysis results reveal that the correlation coefficient of 60 relationships is statistically significant (p < 0.01). The largest Z-values are satisfaction on continuance intention (75.695), confirmation on perceived usefulness (55.921), confirmation on satisfaction (48.176), attitude on continuance intention (35.602), perceived usefulness on continuance intention (34.287), and perceived usefulness on satisfaction (33.995). 2.4.2. Weight analysis This method is used to estimate the importance of a predictor (i.e. independent construct) and predicts the strength of an independent construct (Jeyaraj et al., 2006). The weights of the 60 most used relationships were examined and are presented in columns 9 to 12 in Table 2.1. The value of weight was computed by dividing the number of statistically significant relationships by the total number of studies used. When the weight is one (1), it shows that the relationship within the variables is significant in all the research, but if the weight is zero (0), it indicates that the relationship is not significant through all the studies examined (Jeyaraj et al., 2006). 18 Doctoral Programme in Information Management Table 2.1 - The most frequently used relationships for meta-analysis and weight-analysis (Ordered by dependent constructs). Nº (1) Independent Constructs (2) Dependent Constructs (3) Meta-analysis Weight analysis Total (4) ∑ Sample (5) AVG of cc (6) Z – value (7) 95% confidence interval (low - high) (8) Nonsignifi cant (9) Signif icant (10) Total (11) Weight (Significant/ Total) (12) 1 Hedonic Value Affective Commitment 5 1266 0.339 12.536 0.198 0.301 0 5 5 1.000 2 Relational Capital 5 1266 0.258 9.381 0.206 0.309 0 5 5 1.000 3 Utilitarian Value 5 1266 0.112 3.997 0.057 0.166 1 4 5 0.800 4 Perceived Ease of Use Attitude 5 1209 0.162 5.676 0.095 0.210 2 3 5 0.600 5 Perceived Usefulness 13 4535 0.408 29.164 0.362 0.414 1 12 13 0.923 6 Satisfaction 6 2580 0.481 26.615 0.451 0.510 0 6 6 1.000 7 Service Quality Confirmation 4 1244 0.339 12.434 0.282 0.394 0 4 4 1.000 8 Continuance Intention Continuance Behaviour 5 1526 0.375 15.385 0.331 0.417 0 5 5 1.000 9 Affective Commitment Continuance Intention 5 1266 0.556 22.284 0.517 0.593 0 5 5 1.000 10 Attitude 14 5657 0.441 35.602 0.409 0.458 0 14 14 1.000 11 Effort Expectancy 3 1075 0.253 8.467 0.196 0.308 1 2 3 0.667 12 Flow 5 1243 0.358 13.207 0.308 0.406 1 4 5 0.800 13 Habit 5 1691 0.255 10.713 0.210 0.299 0 5 5 1.000 14 Hedonic Outcome Expectations 7 1951 0.437 20.679 0.435 0.567 0 7 7 1.000 15 Intrinsic Motivation 3 508 0.453 10.977 0.381 0.520 0 3 3 1.000 16 Perceived Behaviour Control 7 3021 0.296 16.763 0.263 0.328 0 7 7 1.000 17 Perceived Ease of Use 6 2280 0.074 3.538 0.013 0.104 2 4 6 0.667 18 Perceived Enjoyment 16 5808 0.187 14.417 0.158 0.210 3 13 16 0.813 19 Perceived Usefulness 41 13686 0.285 34.287 0.265 0.297 4 37 41 0.902 20 Performance 5 3707 0.241 14.962 0.210 0.271 1 4 5 0.800 21 Satisfaction 74 29220 0.416 75.695 0.399 0.419 2 72 74 0.973 22 Subjective Norm 11 4379 0.179 11.970 0.150 0.208 1 10 11 0.909 23 Trust 7 2353 0.239 11.814 0.175 0.260 1 6 7 0.857 24 Utilitarian Value 9 498 0.242 5.493 0.157 0.323 0 9 9 1.000 25 Perceived Usefulness Disconfirmation 3 1149 0.133 4.529 0.076 0.189 0 3 3 1.000 26 Satisfaction Habit 3 755 0.456 13.499 0.397 0.511 0 3 3 1.000 27 Confirmation Perceived Ease of Use 6 1511 0.458 19.214 0.471 0.548 0 6 6 1.000 28 Context 3 5121 0.233 16.981 0.207 0.259 0 3 3 1.000 29 Individualism 3 5121 0.300 22.143 0.275 0.325 0 3 3 1.000 30 Time perception 3 5121 0.177 12.797 0.150 0.203 0 3 3 1.000 31 Uncertainty Avoidance 3 5121 -0.137 -9.863 -0.164 -0.110 0 3 3 1.000 32 Confirmation Perceived Enjoyment 7 2145 0.622 33.705 0.607 0.663 0 7 7 1.000 33 Context 3 5121 0.157 11.325 0.130 0.184 0 3 3 1.000 34 Individualism 3 5121 0.223 16.226 0.197 0.249 0 3 3 1.000 35 Uncertainty Avoidance 3 5121 -0.137 -9.863 -0.164 -0.110 1 2 3 0.667 36 Context Perceived Monetary Value 3 5121 0.200 14.504 0.174 0.226 0 3 3 1.000 37 Individualism 3 5121 0.323 23.965 0.298 0.347 0 3 3 1.000 38 Time Perception 3 5121 0.190 13.760 0.163 0.216 0 3 3 1.000 39 Uncertainty Avoidance 3 5121 -0.233 -16.981 -0.259 -0.207 0 3 3 1.000 40 Confirmation Perceived Usefulness 33 10168 0.504 55.921 0.492 0.522 0 33 33 1.000 41 Context 3 5121 0.223 16.226 0.197 0.249 0 3 3 1.000 42 Disconfirmation 6 1825 0.555 26.703 0.593 0.652 0 6 6 1.000 43 Individualism 3 5121 0.287 21.125 0.262 0.312 0 3 3 1.000 44 Perceived Ease of Use 12 3532 0.327 20.166 0.268 0.329 1 11 12 0.917 45 Time Perception 3 5121 0.177 12.797 0.150 0.203 0 3 3 1.000 46 Uncertainty Avoidance 3 5121 -0.170 -12.281 -0.196 -0.143 0 3 3 1.000 47 Confirmation Satisfaction 35 10918 0.431 48.176 0.395 0.427 0 35 35 1.000 48 Disconfirmation 9 2576 0.576 33.299 0.550 0.601 0 9 9 1.000 49 Hedonic Benefit 3 928 0.349 11.080 0.291 0.404 0 3 3 1.000 50 Information Quality 5 1735 0.248 10.541 0.203 0.292 1 4 5 0.800 51 Perceived Ease of Use 10 7112 0.188 16.042 0.148 0.194 0 10 10 1.000 52 Perceived Enjoyment 12 8018 0.251 22.962 0.230 0.271 2 10 12 0.833 53 Perceived Usefulness 46 18018 0.248 33.995 0.217 0.245 8 38 46 0.826 54 Perceived Value 3 1158 0.203 6.996 0.147 0.258 0 3 3 1.000 55 Performance 3 2579 0.390 20.901 0.357 0.422 0 3 3 1.000 56 Service Quality 5 2608 0.296 15.574 0.058 0.151 2 3 5 0.600 57 Social Benefit 3 928 0.334 10.563 0.160 0.282 0 3 3 1.000 58 System Quality 7 2477 0.279 14.255 0.242 0.315 0 7 7 1.000 59 Trust 3 880 0.313 9.591 0.252 0.371 0 3 3 1.000 60 Utilitarian Benefit 3 928 0.182 5.598 0.119 0.244 0 3 3 1.000 Note: The highlighted relationships are the best predictors of the weight analysis. AVG of cc = average of the correlation coefficient; Z-value = normal standard deviation. 19 Doctoral Programme in Information Management In order to identify the most effective predictors to use IS continuance intention, (Jeyaraj et al., 2006) classified independent variables in two ways: the variables that were evaluated 5 (five) or more times were classified as well-utilised, and the variables evaluated less than 5 (five) times seen as experimental. Additional definitions of Jeyaraj et al., (2006) have been taken into consideration : best predictors – are the relationships that were classified as well-utilised with the weight greater than or equal to 0.8; and promising predictors – relationships that were classified as experimental with the weight equal to 1. The outcomes of the 60 relationships assessed in the weight analysis show that 34 were classified as well-utilised, and 31 as best predictors (highlighted relationships) of the continuance intention to use an IS. Additionally, 24 out of 26 experimental relationships were classified as promising predictors, requiring more evaluation to succeed as best predictors. For future research, we encourage researchers to evaluate such promising predictors. However, in all the studies, no type of relationship was found to be not significant. According to the findings of metaand weight analyses, the most often used dependent variables were continuance intention, satisfaction, and perceived usefulness. Therefore, the most used independent variables to explain continuance intention (used more than ten times) were satisfaction, perceived usefulness, perceived enjoyment, attitude, and subjective norms. 2.4.3. Moderator analysis The analysis of potential moderators was performed using factors from the methodological, economic, and cultural context. Thus, these factors were tested in the relationships that predict continuance intention, and present a considerable number of observations (at least 30) (Geyskens et al., 2009; Lipsey & Wilson, 2001). However, the relationships selected were perceived usefulness and satisfaction to continuance intention (Schmidt & Hunter, 2016). The results of the analysis are presented in Table 2.2. The moderator sample size had a significant moderating effect on the relationship of perceived usefulness to continuance intention ( β = 0.338, M_Low = 0.418, M_High = 20 Doctoral Programme in Information Management 0.316, p < 0.05), and had no significant moderating effect for the other relationships of satisfaction to continuance intention ( β = 0.502, M_Low = 0.461, M_High = 0.448). In the economic context, the moderator economic development had no significant moderating effect of perceived usefulness to continuance intention ( β = 0.467, M_Low = 0.364, M_High = 0.380), and satisfaction to continuance intention ( β = 0.546, M_Low = 0.451, M_High = 0.466). Similarly, the moderator innovation level had no significant moderating effect of perceived usefulness to continuance intention ( β = 0.490, M_Low = 0.330, M_High = 0.414), and satisfaction to continuance intention ( β = 0.491, M_Low = 0.492, M_High = 0.438). In the cultural context, the power distance moderator had a significant moderating effect of satisfaction to continuance intention ( β = 0.455, M_Low = 0.495, M_High = 0.414, p < 0.05), and had no significant moderating effect on the other relationship of perceived usefulness to continuance intention ( β = 0.497, M_Low = 0.320, M_High = 0.445). The moderator individualism had a significant moderating effect of perceived usefulness to continuance intention ( β = 0.499, M_Low = 0.308, M_High = 0.420, p < 0.1), and had no significant moderating effect on the other relationship, satisfaction to continuance intention ( β = 0.526, M_Low = 0.393, M_High = 0.449). The moderator masculinity had a significant moderating effect of satisfaction to continuance intention ( β = 0.527, M_Low = 0.374, M_High = 0.465, p < 0.1) and had no significant moderating effect on the other relationship of perceived usefulness to continuance intention ( β = 0.511, M_Low = 0.336, M_High = 0.435). The moderator uncertainty avoidance had a significant moderating effect on perceived usefulness to continuance intention ( β = 0.361, M_Low = 0.441, M_High = 0.327, p < 0.1), and had no significant moderating effect on the relationship satisfaction to continuance intention ( β = 0.473, M_Low = 0.433, M_High = 0.430). Similarly, the moderator long term orientation had a significant moderating effect on perceived usefulness to continuance intention ( β = 0.332, M_Low = 0.420, M_High = 0.308, p < 0.1), and had no significant moderating effect on the relationship satisfaction to continuance intention ( β = 0.540, M_Low = 0.453, M_High = 0.455). Finally, the moderator indulgence had a significant moderating effect on the relationship satisfaction to continuance intention ( β = 0.584, M_Low = 0.425, M_High = 21 Doctoral Programme in Information Management 0.484, p < 0.1), and had no significant moderating effect on the relationship perceived usefulness to continuance intention ( β = 0.394, M_Low = 0.408, M_High = 0.343). Table 2.2 - Moderation’s analysis. Moderator Level Perceived Usefulness to Continuance Intention Satisfaction to Continuance Intention ! R P_value ! R P_value Sample size Intercept .338 .001 .502 .001 High 1 .316 1 .448 Low .159 .418 .038** .049 .461 .409 Economic development Intercept .467 .001 .546 .001 High 1 .380 1 .466 Low -.067 .364 .454 -.026 .451 .712 Innovation Level Intercept .490 .001 .491 .001 High 1 .414 1 438 Low -.125 .330 .163 .094 .492 .113 Power Distance Intercept .497 .001 .455 .001 High 1 .445 1 .414 Low -.125 .320 .158 .143 .495 .015** Individualism Intercept .499 .001 .526 .001 High 1 .420 1 .449 Low -.166 .308 .055* .015 .393 .836 Masculinity Intercept .511 .001 .527 .001 High 1 .435 1 .465 Low -.138 .336 .153 -.118 .374 .060* Uncertainty avoidance Intercept .361 .001 .473 .001 High 1 .327 1 .430 Low .159 .441 .096* .018 .433 .772 Long Term orientation Intercept .332 .001 .540 .001 High 1 .308 1 .455 Low 166 .420 .055* -.020 .453 .769 Indulgence Intercept .394 .001 .584 .001 High 1 .343 1 .484 Low .063 .408 .431 -.115 .425 .053* Note: *** p < 0.01; ** p < 0.05; * p < 0.10 2.5. Discussion Considering the number of studies on continuance intention to use an IS using theories or models, it becomes significant and suitable to analyse and discuss their collective findings. We can verify that the variables and relationships used are quite dispersed, as studies are analysing different IS technologies, studies from separate times, and different geographical spaces with distinct cultures. The results reveal that the metaand weight analyses for the independent variables on equivalent dependent variables are closer. The higher the weight of an independent variable, the greater is the probability that it is significant in performing the metaanalysis (Rana et al., 2015). In the meta-analysis, all the 31 best predictors, and all the 22 Doctoral Programme in Information Management 24 promising predictors were found to be statistically significant. The remaining three well-utilised relationships, namely service quality on satisfaction, perceived ease of use on attitude, and perceived ease of use on continuance intention, were also statistically significant. The results reveal that the most important relationships to predict continuance intention to use an IS are simultaneously statistically significant in metaanalysis and best predictors in weight analysis. Additionally, the width of the confidence interval depends on the correctness of individual studies along with the number of the cumulative studies (Rana et al., 2015). We can verify that all of the best predictor and promising predictor relationships obtained a narrow interval, providing confidence to the level of variance of the correlation values. According to the results, three theoretical models (Figure 2.3) were designed to support future studies on continuance intention to use an IS. The first model (A) was created using all the data from our analysis (general model), and then, to understand the model's evolution, the data were divided into two groups (from 2001 to 2010, and from 2011 to 2017), and with this information, two more models were created. The second model (B) was created using data from 2001 to 2010, and the last model (C) was created using data from 2011 to 2017. To generate the theoretical models, first, significant relationships from the meta-analysis were selected, second, the best predictors of weight analysis were selected, and finally, direct or indirect variables related to continuance intention to use an IS were selected. Moreover, the relationships such as service quality on satisfaction, perceived ease of use on attitude, and perceived ease of use on continuance intention were evaluated five or more times and with weight less than 0.8. Basically, because they were statistically non-significant in individual studies, future research is needed to prove or disapprove the existing trend (Jeyaraj et al., 2006). According to the findings, it is possible to understand that there is an evolution of the models, just by comparing the original model of IS continuance intention (A Bhattacherjee, 2001) with our general model. The general model is more complex and presents more relationships with a significant set of constructs. Going to model B, it is simpler than the general model (A). This facet means that, in that period (2001 to 2010), the key factors that influenced users to continue using IS were: satisfaction, perceived 23 Doctoral Programme in Information Management usefulness, and attitude. Continuance intention was used to predict continuance behaviour. In model B, Confirmation only explains satisfaction and perceived usefulness over time, and it starts explaining perceived ease of use and perceived enjoyment. This phenomenon also happened with perceived ease of use. It only explains satisfaction, over time, and explains perceived usefulness. In model C, which is more complex than model B and very similar to the general model (A), we can argue that in that period (2011 to 2017), (i) the type of IS increased considerably, then, to support this phenomenon, (ii) the number of constructs also increased. In model B, we have disconfirmation explaining satisfaction and perceived usefulness, but we do not have these relationships in model C. This means that these relationships have been so well-explored that they became part of a collective body of knowledge. In model B, we have perceived ease of use explaining satisfaction; over time, the same construct started explaining perceived usefulness and began to be explained by confirmation. In model C, we do not have continuance intention explaining continuance behaviour. From 2011 to 2017, it is possible to verify that the independent constructs increased significantly. Some constructs belong to other theoretical models, for example, attitude, subjective norms, and perceived behaviour control (Ajzen, 1985), system quality and information quality (DeLone & McLean, 1992) perceived ease of use (Davis, 1989). Two constructs do not appear in either model B or C but appear in the general model (A), such as performance and perceived behaviour control. This element means that these relationships have been explored over the whole range of our analysis (2001 - 2017). Moreover, when the dataset was divided, they did not become statistically significant in both groups. However, over time, the theoretical models grow more complex, and new constructs appear, such as perceived enjoyment, trust, flow, subject norms, and performance, etc. Nevertheless, some constructs are in decadence because they turned into common knowledge, and some constructs like satisfaction, perceived usefulness, confirmation, perceived ease of use, and attitude remain over time. This characteristic means that these constructs are still essential to predict IS continuance intention. Now we can argue that we have a powerful model (A) to predict continuance intention to use IS, using 24 Doctoral Programme in Information Management distinct types of technologies, in different environments. In addition, the theoretical models had a significant evolution. The dashed relationships mean that they were not statistically significant and best predictors in that period. Moreover, the performance and perceived behaviour control on continuance intention were statistically significant and best predictors only with all the dataset (model A). 25 Doctoral Programme in Information Management Figure 2.3 - Theoretical models based on metaand weight analysis (*** p < 0.01; ** p < 0.05; * p < 0.10). Note: The dashed relationships represented above were not statistically significant and best predictors in the period of analysis. 32 Doctoral Programme in Information Management The results underscore that small sample size, high levels of individualism, low levels of uncertainty avoidance, and low levels of long-term orientation positively moderate the relationship between perceived usefulness and continuance intention. Regarding individualism, managers should focus on the task and autonomy of the individual. Individual management is more important than group management (Hofstede & Minkov, 2010). For uncertainty avoidance, the results are important for managers to promote the usability of the IS, considering that they are more receptive to new technology (Baptista & Oliveira, 2015). For long term orientation, the results are pivotal for managers because people with a low level of long-term orientation follow instructions, and respect rules and traditions. However, the results determine that a low level of power distance, a high level of masculinity, and a high level of indulgence positively moderate the relationship between satisfaction and continuance intention. Concerning power distance, independence is a cardinal characteristic, and power should be distributed unequally (Baptista & Oliveira, 2015; Hofstede & Minkov, 2010). For masculinity, managers should promote competitions and training to motivate people. For indulgence, impulses and desires are more important. Managers should satisfy people according to their desires (Hofstede & Minkov, 2010). 2.5.4. Limitations and recommendation The research contains several limitations. First, we did not include certain studies because of the unavailability of their quantitative data, or because they were qualitative. For example, some studies did not report the effect size when the relationships were statistically non-significant. Integrating these studies could generate relevant information in terms of the significance of the constructs. Secondly, in the previous research if the data are biased on IS continuance intention, then the effect of the mean presented through meta-analysis will also reflect this bias, and in this study, the handling publication biases are missing. Moreover, we used meta-essentials, a metaanalysis tool. This tool has limitations, considering that meta-essentials is not able to perform more advanced analyses, like linear model or structured equation model (Rhee et al., 2015), future research should consider a more advanced tool to provide more insights and a different approach to research. 33 Doctoral Programme in Information Management Future research should consider relationships like service quality on confirmation, hedonic outcome expectation, and intrinsic motivation on continuance intention, satisfaction on habit, trust, hedonic benefits, and performance on satisfaction as they were classified as promising predictors and were statistically significant with a high average correlation coefficient. Culture has an important impact on IS continuance intention, and future research should incorporate a culture variable such as subjective norms, habit, or attitude, to provide additional insights. Considering that some IS such as mobile payments and gamification are growing exponentially, future research could contemplate a meta-analysis to synthesise and provide more findings in these subjects. 2.6. Conclusions The goal of our research was to collect and analyse different studies on continuance intention to use IS, by using meta-analysis combined with weight analysis. The systematic review of existing studies comprised 115 papers, which constituted the basis for the analysis of our research. The most used technology was e-learning, followed by social network services. In the process of collecting data for our analysis, we found more than 600 relationships and selected the relationships that had been examined three or more times, totalling 60 different relationships. 34 out of the 60 most used relationships have been examined five or more times. We identified the most used variables in the literature and highlighted their relevance, contributing to the current state of the art (Hamari & Keronen, 2017; K. Wu et al., 2011; Y. Zhao et al., 2018). Concerning moderation analysis, we presented noteworthy results. The relationship perceived usefulness to continuance intention was moderated with sample size, individualism, uncertainty avoidance and long term orientation. Furthermore, the relationship satisfaction to continuance intention was moderated with power distance, masculinity and indulgence. According to our empirical result, the evolution of the theoretical models was presented, using the best predictors and the significant relationships to predict continuance intention to use an IS. This study works as a reference basis for future research that seek to develop the area of continuance intention to use any type of IS. Moreover, among the most used constructs, we found constructs 34 Doctoral Programme in Information Management from the expectation confirmation model (A Bhattacherjee, 2001), technology acceptance model (Davis, 1989), the theory of planned behaviour (Ajzen, 1985), and the DeLone and McLean IS success model (DeLone & McLean, 1992). According to Jeyaraj et al., (2006) criteria, 31 relationships were classified as best predictors (examined five or more times, and weight ≥ 0.8), and 24 relationships were classified as promising predictors (examined fewer than five times, and weight = 1), needing more tests to qualify as the best predictors. 35 Doctoral Programme in Information Management 36 Doctoral Programme in Information Management Chapter 3 – Understanding the factors of mobile payment continuance intention: empirical test in an African context 3.1. Introduction In recent years the number of mobile phone users has been growing exponentially, motivating companies to deliver services via mobile phones (Karjaluoto et al., 2019; Persaud and Azhar, 2012; J. Wu et al., 2017). Mobile payment (m-payment) is one of the many services that can be used via mobile phones. M-payment is a imbursement method that uses mobile phones to make financial transactions such as paying for goods or services, transferring money, and withdrawing money (Fan et al., 2018a; T. Zhou, 2013). M-payment technology was a disruptive revolution that affected payment ecosystems, originating in the United States and spreading throughout the world (Fan et al., 2018a). In Africa, m-payment was launched in Kenya, and was quickly adopted in other countries. Mozambique is one of the African countries that adopted mpayment, helping rural people who do not have banking infrastructure near them (Batista and Vicente, 2018). When we compare the usefulness of m-payment in developed and developing economies, it appears that the impact on people's lives is most noticeable in developing economies, considering that financial services do not reach most of the population yet, and most people travel long distances to access them (Asamoah et al., 2020; Humbani and Wiese, 2018; Iman, 2018). M-payment has a major impact on these communities because it provides basic financial services such as money transfer, payment of goods and services, and/or withdrawing money, thereby improving people's lives (Iman, 2018; Rahman et al., 2020). Much research has been undertaken to understand m-payment in different contexts (e.g., Lu et al. (2017) in China; Lin et al. (2017) in the United States; Sinha et al. (2019) in India; Oliveira et al. (2016) in Portugal). However, few studies have been conducted in the African context (Chen and Li, 2017; Lin et al., 2017). Previous literature has used different theoretical models to understand continuance intention to use m-payment. Shao et al. (2019) used trust and innovation diffusion theory; Lu et al. (2017) used expectation-confirmation theory, mobility, privacy protection, social influence, and cultural values; Chen and Li (2017) used IT continuance, risk-trust, and affect- 37 Doctoral Programme in Information Management cognition literature. With the exception of two studies that integrated quality factors to understand continuance intention of m-payment (T. Zhou, 2013, 2014b), our study shows how important it is to combine DeLone and McLean information system (D&M IS) success model, and expectation-confirmation model (ECM). Each model has strengths and weaknesses, and these are offset and complemented by combining these two models. Despite the benefits of m-payment, there are still barriers to the continuance use. Many users remain concerned about individual performance and the quality of services, since the m-payment service involves transaction information that affects user privacy. It is important that the users feel confident about m-payment, realize that the service is of quality, that it contains useful information and that they feel the need to use more and more. Earlier studies addressed similar issues (Fan et al., 2018a; Shao et al., 2019; T. Zhou, 2011b, 2013), but did not integrate a model that can explain different qualities of mpayment, satisfaction, and perceived individual performance to understand continuance intention of m-payment. Considering the impact of m-payment in the African context, due to the lack of access to technology in the same proportion, a service that can be used anytime and anywhere, regardless of the education and economic level, reducing the need to use banks is of great importance (Pal et al., 2020). The context can challenge the theoretical models to explain m-payment. Based on these reasons, we approached the research question (RQ): How do the individual performance drivers influence the continuance intention to use m–payment in an African context? In this sense, it is in our interest to understand the effects of individual performance drivers combined with the ECM on m-payment. We joined two well-established models, the D&M IS success model (DeLone and Mclean, 2003) and ECM (A Bhattacherjee, 2001) to investigate continuance intention to use m-payment and gain a holistic view of the quality of service and individual performance on continuance intention. The current research contributes to the literature firstly by combining the D&M IS success model with the ECM model with the aim of improving the understanding of continuance intention to use m–payment, identifying important determinants. As per previous research, this would be the first study to combine all the factors of D&M IS success model and ECM model with the purpose to understand continuance intention 38 Doctoral Programme in Information Management to use m–payment. Secondly, considering that the African market is developing, this research will benefit people and companies that are developing IT related to m-payment by identifying the most important factors that can lead to end-user’s long-term usage. Thirdly, by addressing the factors of individual’s continuance intention to use mpayment, the study deepens knowledge, about what is important to the long–term usage of an IS (A Bhattacherjee, 2001). The next section presents the bibliographic review. Section 3 outlines the hypotheses and the research model. Section 4 describes the research methodology. Data analyses and results of research are presented in Section 5. Finally, the discussion and conclusions are detailed in Sections 6 and 7, respectively. 3.2. Literature review 3.2.1. Mobile Payment M-payment is a type of payment that can be performed by mobile devices (such as mobile phones, smartphones, etc.), to pay for goods, services, and bills. They use wireless technologies (mobile phone networks, NFI, Bluetooth, RFID, etc.) to perform transactions (Kaur et al., 2020; Kujala et al., 2017; Liébana-cabanillas et al., 2014, 2018; F. Liébana-Cabanillas and Lara-Rubio, 2017; J. K. Park et al., 2019). Other authors refer to m-payment as a service to carry out payment, check balances, and transfer money in a simple way, anytime and anywhere (Pal et al., 2020; T. Zhou, 2013, 2014b). There are different ways to conduct a transaction using m-payment, the simplest is based on using short-messages with a simple mobile phone whereby the user can check balances or conduct payments using short messages (Luna et al., 2019; Zhou, 2013). Another method is by using NFC (near field communication), communication is established by the proximity of two devices, and the transaction is made (de Luna et al., 2019; Kujala et al., 2017; Francisco Liébana-Cabanillas et al., 2019). The most sophisticated means is by using a mobile application (app), the user downloads the app, installs it on a smartphone, and registers to start using the app (Singh et al., 2020; Verkijika, 2020). 39 Doctoral Programme in Information Management M-payment continuance intention was studied by Lu et al. (2017), who applied mobility, privacy protection, and social influence, and concluded that post-usage privacy protection and social influence belief impact users’ intentions to continue using m-payment. Yu et al. (2018) applied trust transfer theory, perceived similarity, and entitativity, and determined that satisfaction is an important predictor influencing mpayment continuance intention, and that the trust transfer process positively influences continuance intention through satisfaction. Zhou (2013) used information systems success and flow theory, and posited that flow, satisfaction, and trust influence continuance intention; and m-payment providers need to offer a quality system, information, and service to guarantee long-term usage. Chen and Li (2017) applied IT continuance, risk-trust, and affect-cognition, and concluded that satisfaction and postadoption perceived usefulness positively influence continuance intention of mpayment. In general, most previous studies focused on trust and satisfaction of the mpayment continuance intention (J. Lu, Wei, et al., 2017; Tam et al., 2020; T. Zhou, 2012, 2014b), but quality and performance also play an important role in m-payment continuance usage. 3.2.2. Mobile payment in an African context In an African context, telecommunication companies (Safaricom) launched m-payment (M-Pesa) in Kenya in 2007 (Jack and Suri, 2011; Wenner et al., 2018). Kenyans consolidated the usefulness of this technology (Omigie et al., 2017; Uwamariya and Loebbecke, 2020). The technology then grew exponentially and spread across the continent, and is now being used in more than five African countries and has more than 29 million active users (Vodafone Group, 2016; Wenner et al., 2018). M-payment was adopted in Mozambique and grew rapidly, considering that this service is an alternative to bank-based systems for the population to access financial services. Using simple short messages, users can perform a transaction to transfer money, and pay for goods and services. This service was introduced by the Mozambican Telecommunication Company MCel (mKesh) and by Vodacom (M-Pesa) (Ortigão et al., 2015). Batista and Vicente (2018) found that m-payment is very important in rural areas to increase financial inclusion. Jack and Suri (2011) argued that m-payment 40 Doctoral Programme in Information Management spread very quickly because it is an alternative banking service and has substantial impact on people in low economic conditions. Tobbin and Kuwornu (2011) argued that perceived ease of use and perceived usefulness are the most important factors of behavioural intention to use m-payment in Ghana. Humbani and Wiese (2018) show that convenience and compatibility positively influence the adoption of m-payment. Additionally, they demonstrated that only gender could moderate the relationship between convenience and the adoption of m-payment. 3.2.3. Theoretical models 3.2.3.1. DeLone and McLean IS success model The DeLone and McLean IS success model has been broadly used to explain individual and organizational performance (DeLone and McLean, 1992). However, in this study, the focus is on the individual level. The D&M IS model explains that (1) both quality system and information quality significantly influence the use of IS and user satisfaction, (2) the use of IS influences the user’s satisfaction and vice versa, (3) both use and satisfaction significantly influence individual performance, and (4) individual performance significantly influences organizational impact. Several studies confirm that this model is powerful to explain individual performance and can be used with other models or variables (Baabdullah et al., 2019; Sharma and Sharma, 2019). Tam and Oliveira (2016) employed it to understand the impact of mobile banking; Hsu et al. (2014) used the model to explain the repurchase intention on online group-buying; Wang (2008) applied it to explain the impact of e-commerce system success. After ten years DeLone and Mclean (2003) reviewed several papers that validate, challenge, and propose improvements to the original model, and proposed an updated model. In the updated model they include the fact that service quality significantly influences the use of the system and satisfaction. They realized that with the growth of IS, users started paying attention to the quality of services (Tam and Oliveira, 2016). Concerning the impact, in the updated model, they realized that other studies have proposed several types of impacts and decided to join all the impacts into a single 41 Doctoral Programme in Information Management impact called net benefits. Considering that the proposed model is based on the individual level, the individual performance will also be used. 3.2.3.2. Information System Continuance Model There are several theories used in studies related to information systems, such as the unified theory of acceptance and usage of technology (UTAUT), the technology acceptance model (TAM), and the task-technology fit (TTF). Our interest is based on the post-adoption theoretical model of IS, at the individual level. There is a difference between adoption and continuance intention of IS. Adoption refers to factors that explain why an individual adopts or rejects a technology (Humbani and Wiese, 2019; Straub, 2009). At this stage, the users have their first contact with the technology, and depending on their experience, they decide whether to use it or to reject it. Continuance intention refers to factors that explain why an individual uses a technology for a long time, thus contributing to the continued use of the technology (Franque et al., 2020; X. Lin et al., 2017). It involves understanding the long-term factors that contribute to the success of the IS (A Bhattacherjee, 2001; X. Lin et al., 2017). Despite the merit of previous studies that applied adoption theories such as UTAUT and TAM to explain continuance intention (Hadji and Degoulet, 2016; Joo et al., 2016; Wu and Chen, 2017), the application of these models may suffer from some limitations, leading to misunderstandings and misapplications of these theories (Bhattacherjee and Barfar, 2011; Franque et al., 2020; Nabavi et al., 2016). The ECM model was based on the expectation-confirmation theory of Oliver (1986). The model explains that (1) continuance intention to use IS was strongly anticipated by user satisfaction, followed by users’ perceived usefulness of the system, (2) user satisfaction was predicted by users’ confirmation of perceived usefulness and expectation, and (3) user confirmation of expectation was a significant predictor of users’ perceived usefulness. The model was extensively tested in IS research and confirmed to be a good model to explain continuance intention (Carillo et al., 2017; Ryu, 2018; Talwar et al., 2020; Wang et al., 2019; Zheng, 2019). Users’ satisfaction is the best factor to improve the continuance 48 Doctoral Programme in Information Management 3 Service Quality (SERQ) SERQ1. The responsible service personnel are always highly willing to help whenever I need support with the M-Payment. SERQ2. The responsible service personnel provide personal attention when I experience problems with the M-Payment. SERQ3. The responsible service personnel provide services related to M-Payment at the promised time. SERQ4. The responsible service personnel have enough knowledge to answer my questions with respect to M-Payment. (Tam & Oliveira, 2016) 4 Use (U) U1. I use M-Payment. U2. I use M-Payment to buy products and services. U3. I use M-Payment to make transfers. U4. I use M-Payment to withdraw money. (Venkatesh, 2003) 5 Satisfaction (S) S1. I am very pleased to use M-Payment. S2. I am very happy with M-Payment. S3. I am delighted with M-Payment. (A Bhattacherjee, 2001) 6 Confirmation (C) C1. My experience with using M-Payment was better than I expected. C2. The service level provided by M-Payment was better than I expected. C3. Overall, most of my expectations from using M-Payment were confirmed. (A Bhattacherjee, 2001) 7 Perceived Usefulness (PU) PU1. Using M-Payment improves my performance. PU2. Using M-Payment increases my productivity. PU3. Using M-Payment enhances my effectiveness. PU4. I find M-Payment to be useful for my work. (A Bhattacherjee, 2001) 8 Individual Performance (IP) IP1: M-Payment enables me to accomplish tasks more quickly. IP2: M-Payment makes it easier to accomplish tasks. IP3: M-Payment is useful for my job. (Tam & Oliveira, 2016) 9 Continuance Intention (CI) CI1. I intend to continue using M-Payment rather than discontinue its use. CI2. My intentions are to continue using M-Payment rather than manual processing or other alternative means. CI3. I plan to continue using M-Payment in my job. (A Bhattacherjee, 2001) 3.4.2. Data Data were collected from June 2018 to October 2018. The items were assessed on a seven-point scale, ranging from one (totally disagree) to seven (totally agree). The question was created and managed in English and revised for content validity by a language expert. Nevertheless, a professional translator translated the questionnaire into Portuguese to adjust it to the Mozambican context. The questionnaire was reverse translated to English by a different translator to ensure equivalence (Brislin, 1970). To validate the instruments, a pilot test was conducted on a group of 40 students, who were excluded from the main sample. Given that the goal of this study is to investigate continuance intention to use m-payment, the target respondents should have experience in using m-payment. To ensure this, the valid respondents in the Mozambican context were confined to M-Pesa and Mkesh users. The survey was sent to the respondents, providing a hyperlink for the questionnaire. We received 338 valid responses by the end of October 2018 from the 900 e-mails sent, which corresponds to a 37.5% response rate. We tested the sample distributed to the first and second respondent groups using 49 Doctoral Programme in Information Management the Kolmogorov-Smirnov (K-S) test and confirmed that they do not differ statistically (Ryans, 1974), showing that non-response bias was not present. The common method bias was also examined using Harman’s test (Podsakoff et al., 2003) confirming no significant common method bias in the data. The characteristics of the sample are shown in Table 3.2; 60% of the respondents were men, 42% of the respondents had used m-payment one (1) to four (4) times during the last 3 months. Table 3.2 - Sample characteristics. Age < 25 119 35% 25 - 30 96 28% 31 - 40 76 22% 41 - 50 38 11% > 50 9 3% Gender Female 134 40% Male 204 60% Education High school or below 83 25% Bachelor’s degree 151 45% Master’s degree or higher 104 31% Employment Students 95 28% Working professionals 203 60% Retired 1 0% Unemployed 39 12% Marital status Single 162 48% Married 74 22% Divorced 23 7% Widowed 10 3% Common-law marriage (cohabitation) 67 20% Do not know answers 2 1% M-payment usage frequency (time / 3 months) 1 - 4 143 42% 5 - 10 92 27% > 10 103 30% 50 Doctoral Programme in Information Management 3.5. Data analysis and results Structural equation modeling (SEM) with partial least square (PLS) was used to test and assess the validity of the theoretical model. Previous research has recognized the potential of SEM for measuring structural models (Alraimi et al., 2015; Tam and Oliveira, 2016). SEM is a set of statistical models used to assess the validity of theories with empirical data (Ringle et al., 2005). Additionally, the Kolmogorov–Smirnov (KS test) was implemented, as it is used in cases where data are not usually distributed, and the research model is complex and has not been tested in the literature. Thus, PLS is the appropriate method for this research. Smart PLS 3 software (Ringle et al., 2015) was used to analyse the theoretical model relationships. 3.5.1. Measurement model The results of the measurement model are presented in Tables 3.3 and 3.4. The results of composite reliability (CR) are greater than 0.70, indicating that the model has good internal consistency. To assess the indicator reliability, we considered a loading greater than 0.70. The instruments present a good indicator reliability. Average variance extracted (AVE) was used to test convergent validity. AVE should be greater than 0.50 so that the latent variables explain more than half of the variance of their indicators (Fornell and Larcker, 1981; Hair Jr et al., 2016; Henseler et al., 2009). 51 Doctoral Programme in Information Management Table 3.3 - Measurement model. Construct AVE Composite Reliability Cronbach's Alpha Item Loadings t-value Information Quality (INFQ) 0.652 0.882 0.822 INFQ1 0.829 38.958 INFQ2 0.851 44.922 INFQ3 0.809 40.380 INFQ4 0.737 17.910 System Quality (SYSQ) 0.628 0.871 0.801 SYSQ1 0.736 17.964 SYSQ2 0.839 32.255 SYSQ3 0.837 42.697 SYSQ4 0.753 20.061 Service Quality (SERQ) 0.617 0.865 0.792 SERQ1 0.725 15.049 SERQ2 0.824 30.969 SERQ3 0.818 31.505 SERQ4 0.771 23.059 Use (U) 0.595 0.855 0.773 U1 0.781 27.522 U2 0.745 21.760 U3 0.792 22.677 U4 0.767 23.880 Satisfaction (S) 0.690 0.870 0.776 S1 0.850 48.015 S2 0.833 40.310 S3 0.809 27.531 Confirmation (C) 0.623 0.832 0.697 C1 0.777 21.874 C2 0.824 32.682 C3 0.765 24.587 Individual performance (IP) 0.630 0.836 0.705 IP1 0.782 27.527 IP2 0.846 39.725 IP3 0.751 18.229 Perceived Usefulness (PU) 0.575 0.844 0.753 PU1 0.729 17.764 PU2 0.813 35.602 PU3 0.778 25.344 PU4 0.709 15.089 Continuance Intention (CI) 0.525 0.812 0.695 CI1 0.757 22.151 CI2 0.769 20.020 CI3 0.807 41.005 Note: AVE: Average variance extracted. 52 Doctoral Programme in Information Management Table 3.4 - Latent construct correlations and square roots of AVEs. Mean STDEV INFQ SYSQ SERQ U S C IP PU CI INFQ 4.869 1.123 0.807 SYSQ 4.813 1.123 0.608 0.793 SERQ 4.247 1.166 0.247 0.284 0.785 U 4.693 1.154 0.509 0.395 0.349 0.771 S 4.464 1.165 0.524 0.432 0.361 0.566 0.831 C 4.330 1.102 0.498 0.413 0.396 0.510 0.640 0.789 IP 4.545 1.055 0.384 0.390 0.257 0.415 0.333 0.397 0.794 PU 4.464 1.085 0.446 0.327 0.329 0.495 0.530 0.566 0.485 0.758 CI 4.481 0.986 0.389 0.342 0.288 0.496 0.426 0.439 0.481 0.386 0.724 Notes: Values in bold are the square root of the average variance extracted; STDEV: Standard deviation; INFQ: Information Quality; SYSQ: System Quality; SERQ: Service Quality; U: Use; S: Satisfaction; C: Confirmation; IP: Individual performance; PU: Perceived Usefulness; CI: Continuance Intention. As seen in Table 3.3, all the constructs meet these criteria, ensuring convergence. This shows that the constructs can be used to assess the theoretical model. Discriminant validity was measured using Fornell-Larcker criterion (Table 3.4), cross-loadings criterion (Table 3.5), and the heterotrait-monotrait ratio of correlations (HTMT) (Table 3.6). Table 3.4 reports the square root of the AVE in bold along the diagonal, and the correlations between the constructs. Based on Fornell and Larcker (1981) criterion, the square root of the AVE should be greater than the correlation between the constructs, and thus the constructs fulfil the criterion. To ensure the discriminant validity, each item presents a higher loading on its corresponding factor than in the cross-loading (Chinn, 1998; Götz et al., 2010). Based on HTMT (Table 3.6) it can be seen that all the values are below 0.90, and it therefore can be concluded that there is discriminant validity (Henseler et al., 2015). The measurement model findings indicate that the model has a good internal consistency, reliability indicator, convergence validity, and discriminant validity, illustrating that the constructs are statistically different and can be used to assess the structural model. 53 Doctoral Programme in Information Management Table 3.5 - Cross loadings. INFQ SYSQ SERQ U S C IP PU CI INFQ1 0.829 0.427 0.185 0.476 0.461 0.428 0.359 0.411 0.329 INFQ2 0.851 0.541 0.166 0.440 0.440 0.385 0.364 0.369 0.354 INFQ3 0.809 0.516 0.222 0.379 0.434 0.420 0.312 0.408 0.327 INFQ4 0.737 0.490 0.234 0.334 0.346 0.371 0.183 0.231 0.235 SYSQ1 0.469 0.736 0.230 0.314 0.328 0.277 0.269 0.199 0.198 SYSQ2 0.477 0.839 0.255 0.306 0.385 0.339 0.326 0.264 0.293 SYSQ3 0.483 0.837 0.242 0.302 0.360 0.343 0.332 0.321 0.312 SYSQ4 0.499 0.753 0.171 0.332 0.293 0.346 0.306 0.247 0.277 SERQ1 0.245 0.278 0.725 0.251 0.289 0.319 0.174 0.275 0.216 SERQ2 0.172 0.208 0.824 0.295 0.283 0.331 0.229 0.271 0.265 SERQ3 0.176 0.168 0.818 0.282 0.287 0.297 0.170 0.232 0.195 SERQ4 0.183 0.239 0.771 0.268 0.275 0.296 0.235 0.253 0.226 U1 0.426 0.325 0.316 0.781 0.402 0.357 0.346 0.353 0.438 U2 0.386 0.292 0.219 0.745 0.406 0.381 0.308 0.421 0.425 U3 0.354 0.248 0.255 0.792 0.436 0.408 0.308 0.337 0.318 U4 0.400 0.346 0.282 0.767 0.498 0.426 0.316 0.414 0.349 S1 0.495 0.415 0.332 0.585 0.850 0.502 0.315 0.445 0.368 S2 0.432 0.348 0.267 0.435 0.833 0.545 0.259 0.469 0.367 S3 0.372 0.307 0.299 0.378 0.809 0.552 0.252 0.407 0.325 C1 0.341 0.294 0.294 0.402 0.561 0.777 0.352 0.456 0.341 C2 0.446 0.369 0.349 0.417 0.497 0.824 0.341 0.456 0.375 C3 0.390 0.312 0.294 0.387 0.454 0.765 0.241 0.426 0.321 IP1 0.309 0.286 0.234 0.356 0.318 0.354 0.782 0.369 0.389 IP2 0.295 0.333 0.181 0.330 0.249 0.278 0.846 0.375 0.403 IP3 0.312 0.309 0.198 0.301 0.223 0.314 0.751 0.414 0.351 PU1 0.388 0.227 0.230 0.347 0.420 0.468 0.322 0.729 0.249 PU2 0.323 0.234 0.260 0.410 0.375 0.466 0.380 0.813 0.325 PU3 0.312 0.283 0.275 0.344 0.431 0.395 0.395 0.778 0.335 PU4 0.333 0.252 0.230 0.402 0.386 0.381 0.377 0.709 0.258 CI1 0.310 0.261 0.231 0.389 0.312 0.339 0.380 0.337 0.757 CI2 0.276 0.271 0.175 0.367 0.334 0.317 0.319 0.274 0.769 CI3 0.359 0.316 0.248 0.431 0.373 0.364 0.462 0.296 0.807 Notes: Indicator loading (in bold) greater than all its cross-loadings; INFQ: Information Quality; SYSQ: System Quality; SERQ: Service Quality; U: Use; S: Satisfaction; C: Confirmation; IP: Individual performance; PU: Perceived Usefulness; CI: Continuance Intention. 54 Doctoral Programme in Information Management Table 3.6 - Heterotrait-monotrait ratio of correlations (HTMT). INFQ SYSQ SERQ U S C IP PU CI INFQ SYSQ 0.755 SERQ 0.311 0.357 U 0.631 0.500 0.444 S 0.648 0.544 0.460 0.721 C 0.657 0.551 0.533 0.694 0.871 IP 0.497 0.519 0.345 0.561 0.447 0.564 PU 0.561 0.422 0.426 0.650 0.695 0.778 0.670 CI 0.487 0.433 0.386 0.658 0.563 0.623 0.656 0.527 Notes: INFQ: Information Quality; SYSQ: System Quality; SERQ: Service Quality; U: Use; S: Satisfaction; C: Confirmation; IP: Individual performance; PU: Perceived Usefulness; CI: Continuance Intention. 3.5.2. Structural model After the validation of the measurement model, the structural model was analysed for hypotheses and constructs testing. Figure 3.2 presents the research results. The structural model assessment used 5000 bootstrap resamples to estimate the path significance level of the model (Henseler et al., 2009). The VIF (variance inflation factor) was also tested to assess the multicollinearity. All of the constructs are below the threshold of 5, indicating the absence of multicollinearity between the constructs (Hair Jr. et al., 2016). Figure 3.2 - Research model. Notes: Integration of Expectation Confirmation Model (ECM) and DeLone and McLean information system success (D&M IS); (*** p < 0.01; ** p < 0.05; * p < 0.10). 55 Doctoral Programme in Information Management The model explains 32% of the variation in the use of m-payment. The information quality ( " # = 0.400, p < 0.01) and service quality ( " # = 0.225, p < 0.01) are statistically significant in explaining use, thus confirming H1a and H3a. The system quality is not statistically significant in explaining use, and thus hypothesis H2a is not confirmed. The model explains 53% of the variation in satisfaction in using m-payment. The information quality ( " # = 0.152, p < 0.01), confirmation ( " # = 0.389, p < 0.01), use ( " # = 0.240, p < 0.01), and perceived usefulness ( " # = 0.136, p < 0.05) are statistically significant in explaining satisfaction, thus confirming H1b, H4a, H5a, and H7a. System quality and service quality are not statistically significant in explaining satisfaction, and thus hypotheses H2b and H3b are not confirmed. The model explains 34% of the variation in confirmation of m-payment. Information quality ( " # = 0.359, p < 0.01) and service quality ( " # = 0.275, p < 0.01) are statistically significant in explaining confirmation, thus confirming H1c and H3c. System quality is not statistically significant, and thus hypothesis H2c is not confirmed. The model explains 32% of variation in perceived usefulness of m-payment. Confirmation ( " # = 0.566, p < 0.01) is statistically significant in explaining perceived usefulness, thus confirming H4b. The model explains 19% of variation in individual performance. Use ( " # = 0.416, p < 0.01) and satisfaction ( " # = 0.146, p < 0.1) are statistically significant in explaining individual performance, therefore confirming H5b and H6a. 36% of the variation is explained by the model in continuance intention to use mpayment. Use ( " # = 0.272, p < 0.01), individual performance ( " # = 0.310, p < 0.01), and satisfaction ( " # = 0.270, p < 0.01) are statistically significant in explaining continuance intention, therefore confirming H5c, H6b, and H8. Perceived usefulness is not statistically significant, and thus hypothesis H7b is not confirmed. 56 Doctoral Programme in Information Management The strongest relationships were confirmation on perceived usefulness ( " # = 0.566), use on individual performance ( " #= 0.416), information quality on use ( " #= 0.400), and confirmation on satisfaction ( " # = 0.389). 3.6. Discussion The model proposed is a combination of the D&M IS success model (DeLone and Mclean, 2003) and the ECM (A Bhattacherjee, 2001), to explain continuance intention to use m-payment. Based on the findings (see Table 3.7), of 19 hypotheses 14 were confirmed and 5 were not. Therefore, we can argue that most of the hypothesized relationships were confirmed. Individual performance is the strongest predictor of continuance intention, followed by use and satisfaction. Information quality and service quality determine confirmation. Information quality, use, confirmation, and perceived usefulness determine users’ satisfaction. Use is explained by information and service quality. Individual performance is explained by use and satisfaction of m-payment. Perceived usefulness is explained by confirmation. Surprisingly, service quality does not explain satisfaction, perceived usefulness does not explain continuance intention, and system quality explains none of the proposed relationships. Table 3.7 - Hypotheses results. Hypothesis Independent Construct → Dependent construct Findings (β) P value Conclusion H1a Information Quality → Use 0.400 0.000 Supported H1b Information Quality → Satisfaction 0.152 0.008 Supported H1c Information Quality → Confirmation 0.359 0.000 Supported H2a System Quality → Use 0.088 0.235 Not supported H2b System Quality → Satisfaction 0.065 0.261 Not supported H2c System Quality → Confirmation 0.116 0.140 Not supported H3a Service Quality → Use 0.225 0.001 Supported H3b Service Quality → Satisfaction 0.067 0.163 Not supported H3c Service Quality → Confirmation 0.275 0.000 Supported H4a Confirmation → Satisfaction 0.389 0.000 Supported H4b Confirmation → Perceived Usefulness 0.566 0.000 Supported H5a Use → Satisfaction 0.240 0.000 Supported H5b Use → Individual performance 0.416 0.000 Supported H5c Use → Continuance Intention 0.272 0.000 Supported 57 Doctoral Programme in Information Management H6a Satisfaction → Individual performance 0.146 0.071 Supported H6b Satisfaction → Continuance Intention 0.162 0.018 Supported H7a Perceived Usefulness → Satisfaction 0.136 0.025 Supported H7b Perceived Usefulness → Continuance Intention 0.016 0.834 Not supported H8 Individual performance → Continuance Intention 0.310 0.000 Supported The results indicate that information quality positively influences use (H1a), satisfaction (H1b), and confirmation (H1c) of m-payment. The results of H1a and H1b are consistent with those of Tam and Oliveira (2016) and Cidral et al. (2018), who assert that information quality positively influences use and satisfaction. This means that when m-payment provides information with quality, it positively impacts the usage, satisfaction, and confirmation of the expectation of m-payment (Chang et al., 2014; Cheng, 2014; Cidral et al., 2018). The users may always expect to access comprehensive, accurate, and up-to-date information on the m-payment system. As users understand that the m-payment presents quality information, they will understand that the m-payment provider is maintaining the information up to date, and they will therefore continue using m-payment. If the information is out of date or inaccurate, it will negatively influence the usage, satisfaction, and expectations (Gao et al., 2015). Our results indicate that system quality did not affect use (H2a), satisfaction (H2b), or confirmation (H2c) of m-payment, indicating that in the post-acceptance phase the quality of the system is not important for the usage, user satisfaction, or the confirmation of the expectations regarding continuance intention of m-payment. This result is not consistent with the findings reported in several studies (Budiardjo et al., 2017; Cidral et al., 2018; Gao et al., 2015). A possible explanation may reside in the fact that users already assume that m-payment works well, and that it is a mature technology. Another reason may be that our study was conducted in a developing economy context, and that the other studies were conducted in developed economies. A third reason may have to do with the technology studied. The other studies applied different technologies such as web-based learning, social cataloguing sites, online learning, etc. (Cidral et al., 2018; Daʇhan and Akkoyunlu, 2016; Gao et al., 2015). Furthermore, our results indicate that service quality positively influenced use (H3a) and confirmation (H3c), but not satisfaction (H3b). The result of H3a is contradictory with that of Cidral et al. (2018) and Tam and Oliveira (2016), where the relationship 64 Doctoral Programme in Information Management Chapter 4 – Continuance intention of mobile payment: TTF model with trust in an African context: case study of Mozambique 4.1. Introduction How do trust and the alignment of task and technology characteristics affect the continuance use of mobile payment (m-payment)? M-payment has become one of the prominent services in use today (Gao & Waechter, 2015; T. Zhou, 2014b), by which users can make payments for goods, services, and bills; check balances; and transfer money anytime and anywhere (Kujala et al., 2017; Liébana-cabanillas et al., 2018; J. Lu, Wei, et al., 2017; Oliveira et al., 2016; T. Zhou, 2014b). M-payment is defined as any payment in which a mobile device (mobile phone or tablet) is used to perform financial value exchanges (initiate, authorize, and confirm) in return for goods and services (Shao et al., 2019). There are different ways to conduct a transaction using mpayment. The simplest way is short-message-based, by which using a simple mobile phone the user can check balances or conduct payment using short text messages (Singh et al., 2020; T. Zhou, 2013). The m-payment service is growing exponentially all over the world and is bringing benefits to users and the providers (Humbani & Wiese, 2018). Considering its benefits, companies are providing it in different ways around the globe (Fan et al., 2018a; Singh et al., 2020). The technological sector in the African context is still under development, which represents a challenge for the entire economy. Financial services like banks are not available in the same proportions in all regions of a country, e.g. Mozambique (Humbani & Wiese, 2018; INE, 2019), so people living in rural areas are forced to travel farther to gain access to banks (Batista & Vicente, 2014; Humbani & Wiese, 2019). Due to the lack of access to technology in the same proportion; a service that can be used by the population, regardless of their literary and economic level, reducing the need to use banks, is of great importance (Pal et al., 2020). For m-payment, the technological factors in the African context facilitate its accessibility (e.g., only need telephone network) (Jack & Suri, 2011), thereby bringing the service to many people. 65 Doctoral Programme in Information Management Much research has been directed to the subject of m-payment (J. Lu, Wei, et al., 2017; Oliveira et al., 2016; Sinha et al., 2019). Previous literature reports the use of different theoretical models to investigate continuance intention to use m-payment. Zhou, (2013a) used information system success and flow theory to examine continuance intention of m-payment services. Shao et al., (2019) used trust and innovation diffusion theory to understand m-payment platforms. Lu et al., (2017) used expectationconfirmation theory, mobility, privacy protection, social influence, and cultural values to understand m-payment continuation. Chen & Li, (2017) used IT continuance, risktrust, and affect-cognition literature to understand continuance intention of m-payment services. To the best of our knowledge, there are no published studies that examine the intention to continue using m-payment in Mozambique. To fill the gap, we investigate the factors that may influence the continued use of m-payment in that country. Considering that after the change from traditional payment method (cash) to mpayment, users may have uncertainty and mistrust when using m-payment. Knowing that the transactions involve cash, as well as the importance of functionalities and technological factors that may better help the use of m-payment, we analyse the tasktechnology fit (TTF) (Goodhue and Thompson, 1995) and overall trust (Oliveira et al., 2017) theories and evaluate their relationships for continuance intention. Furthermore, no previous study has joined the TTF, overall trust, and ECM constructs into a single model, as we do in the present study (Davison and Martinsons, 2016). The TTF model states that IT can have a positive impact on an individual's task performance if the IT functionalities match the requirements of the tasks that the user needs to perform (Goodhue and Thompson, 1995). Regarding the TTF, the model was used in different technological contexts such as: mobile banking (Tam and Oliveira, 2016), MOOCs (Wu and Chen, 2017), social media search system (Dang et al., 2020), and internet banking (Rahi et al., 2020), and has been combined with different theoretical models such as: technology continuance theory (Rahi et al., 2020), technology acceptance model (TAM) (Wu and Chen, 2017), DeLone and McLean information systems success model (D&M ISSM) (Tam and Oliveira, 2016), mental workload and unified theory of acceptance and use of technology (UTAUT) (Dang et al., 2020). Overall trust refers to the combination of factors such as competence, integrity, and benevolence in order to understand the user’s confidence to use IT (Oliveira et al., 2017). Many studies have 66 Doctoral Programme in Information Management used overall trust in combinations with different models and in different contexts. Oliveira et al. (2017) used overall trust with consumer characteristics, firms’ characteristics, website infrastructure, and interactions to examine purchase intention in e-commerce. Tam et al. (2019) used overall trust with D&M ISSM to examine individual performance in e-commerce. Considering that cyber security is today considered a fundamental challenge to any country or organization, and especially in mobile payments contexts, it is important to have a sense of security and trust. Considering that fact, we expect that there is a positive link between overall trust and use and continuance intention. In addition to trust, the asymmetry of information and communication technology (ICT) across African countries is another challenge. The constraints of that asymmetry could affect the long-term of viability of mobile payment, and for that reason it would be valuable to understand how the task-technology-fit of mobile payment may explain the use and individual performance in Mozambique. The contribution of this study is twofold. First, it contributes to the literature on continuance intention, since many studies have been carried out in the context of technology adoption (Khalilzadeh et al., 2017; Verkijika, 2020). However, studying continuance intention has an impact on the long-term survival of technology (A Bhattacherjee, 2001). For this reason, this study aims to expand the knowledge on this topic by joining the two theories mentioned above: task-technology characteristics and expectation-confirmation model. Second, this study was based on the ECM model (A Bhattacherjee, 2001), which has been tested in different contexts such as m-payment platforms (Shao et al., 2019), MOOCs (Gao et al., 2015) and mobile apps (Tam et al., 2020). However, information systems (IS) have different characteristics and functionalities (Nascimento et al., 2018). In this sense, we seek to extend the model to understand the impact of technological aspects (TTF) and trust in continuance intention to use m-payment. In addition, we intend to understand the impact of satisfaction as a moderator. To the best of our knowledge, this study is the first to combine the TTF, ECM, and the trust dimension to investigate continuance intention. The plan of the paper is as follows. We begin with a literature review of the relevant studies regarding m-payment, TTF, and the dimension of trust. Second, we present the 67 Doctoral Programme in Information Management research model, followed by the hypotheses. Third, we present the methodologies used to test the hypotheses. We then show the results, followed by a discussion and implications of this study, and suggestions for future research. 4.2. Literature review 4.2.1. Mobile Payment in an African context In Africa m-payment has over 29 million active users in more than five countries (Jack and Suri, 2011; Vodafone Group, 2016). In some countries it is called mobile money (Koloseni and Mandari, 2017). M-payment refers to services that enable users to transfer money, pay services and goods, and withdraw money via a mobile phone (Koloseni and Mandari, 2017; Shao et al., 2019). It is sometimes confused with online payment, but it is not the same because online payment uses any mobile device connected to the internet, such as a tablet, mobile phone, or laptop (T. Zhou, 2015), while m-payment uses only mobile phones with or without internet. In Mozambique 26.4% of the population uses a mobile phone and 11.7% uses m-payment (INE, 2019). This exponential growth is occurring because this service is becoming an alternative solution for rural and urban people to access financial services (Humbani and Wiese, 2019). As most bank branches are distant from people, they naturally wish to avoid travelling long distances to access the bank’s services. M-payment thus offers substantial benefits by having an appropriate account and being able to use it anytime and anywhere (Jack and Suri, 2011). Earlier studies have pointed out that m-payment originated in developing countries via SMSs and that it spread quickly due to limited cash alternatives, such as bank accounts and credit cards, thereby helping communities that were otherwise excluded from the financial system (Humbani and Wiese, 2019; Makina, 2017). The value of conducting studies in the African context is therefore evident (Humbani and Wiese, 2019). Once a new technology such as m-payment is developed and released to the society, it is important to understand the acceptance factors (Venkatesh et al., 2003). When the technology has been in use for some time, it is important to investigate the factors that influence the continuity of use from the users' point of view (A Bhattacherjee, 2001). 68 Doctoral Programme in Information Management Regarding m-payment, considering the impact of mistrust or uncertainty, the functionalities, and the technological drivers for m-payment use, it is of great importance and necessity to examine the users' continued use intention and perceptions toward m-payment. Furthermore, it is essential to offer effective m-payment functionalities that can better handle user transactions (T. Zhou, 2014b). Therefore, we decided to use TTF and overall trust as our base theories and examine their impact on the continuance intention. Earlier studies on continuance intention to use m-payment have been published (X. Chen and Li, 2017; Lu et al., 2017; Park et al., 2017; Yu et al., 2018). Considering that our aim is to understand the intention to continue using m-payment, we reviewed previous studies to understand what has already been done in the context of m-payment. Zhou (2014) investigated the continued use of m-payment based on trust, flow, system quality, and information quality. Chen and Li (2017) investigated the intention to continue to use m-payment services using IT continuance, risk-trust, and affectcognition. Koloseni and Mandari (2017) used the theory of planned behaviour, perceived cost, perceived trust, and satisfaction to examine continuance usage of mobile money services. Yu et al. (2018) investigated the intention to continue using mpayment based on trust transfer theory. Humbani and Wiese (2019) used the technology readiness index to predict adoption and continuance intention, with the goal of exploring the readiness of the m-payment app technology. Shao et al. (2019) investigated antecedents of trust and continuance intention in m-payment platforms based on trust and innovation diffusion theory. Raman and Aashish (2021) investigated the antecedents of users' willingness to continue using m-payment services, based on trust, convenience, social value, satisfaction, service quality, attitude, risk, and effort expectancy. Odoom and Kosiba (2020) used UTAUT to investigate continuance intention of m-money. Liébana-Cabanillas et al. (2021) investigated the determinants of intention to continue using and the moderating effect of gender and age of NFC mpayment users. They applied constructs from different models such as theory of reasoned action, perceived value theory, UTAUT2, personal innovation in information technology, mobile payment technology acceptance model, and ECM. With our analysis of the literature, we found that there are different models applied to study mpayment. Regarding the ICT asymmetry across African countries and general 69 Doctoral Programme in Information Management perception of trust to give transparency and security to mobile payment, we joined TTF and overall trust to the ECM. 4.2.2. Information system continuance model “IS continuance” was presented by Bhattacherjee (2001) to explain the intention to continue using IS. The ECM model focuses on three cognitive feelings (expectation of confirmation, perceived usefulness, and satisfaction). The model proposes that satisfaction is the strongest influencer of continuance intention, on the grounds that it results from the confirmation of expectations and the perception of performance. This means that after the user adopts the IS, and uses it for a while, (s)he will realize if the expectations have been confirmed and if the IS is/are useful. If the result is positive, the level of satisfaction increases and, consequently, the intention to continue using the IS increases (A Bhattacherjee, 2001). The model has been used in previous studies, has been integrated with other models, and has been applied in different contexts (Franque et al., 2020). Wu & Chen (2017) integrated TAM, TTF, MOOCs features and social motivation to understand continuance intention to use MOOCs. Gao et al. (2015) joined the information success model, flow theory, and trust to investigate mobile purchase. Humbani & Wiese (2019) integrated the technology readiness index to explore the use of mobile apps. In the present study we combine TTF and overall trust in order to understand m-payment. The model will help us to assess the effect of TTF and overall trust factors on continuance intention. 4.2.3. Task technology fit (TTF) model The TTF model introduced by Goodhue & Thompson (1995) is applied in IS research to explain the performance impact of IS. The theory argues that the ability of an IS to perform an activity task characteristic easily and well, and the technology characteristics that support such activities, significantly influence the use and performance impact of an IS. When the task and technology fit together, the users’ activities are facilitated, improving the use of the technology, and thereby improving the perception of performance impact (Goodhue & Thompson, 1995). The better alignment of the task and technology characteristics makes it possible to encourage the 70 Doctoral Programme in Information Management use and performance impact of m-payment. Several research papers apply the TTF model combined with other theories. Oliveira et al. (2014) combined TTF, UTAUT, and initial trust model (ITM) to investigate mobile banking adoption. The authors used TTF to predict the performance expectancy and adoption in the UTAUT model. Tam & Oliveira (2016b) combined TTF and the IS Success model to understand the influence of mobile banking on individual performance. They used TTF to predict usage and individual performance (performance impact) constructs and as a moderator of the relationship between user satisfaction and individual performance. Dang et al. (2018) combined mental workload (MWL), TTF, and UTAUT to examine the impacts of mental workload and TTF on acceptance of the social media search system. They used TTF to predict the performance expectancy, effort expectancy, and facilitating conditions of the original UTAUT model. Wu & Chen (2017) combined TTF and TAM to understand continuance intention to use MOOCs, using TTF to predict the constructs’ perceived usefulness and perceived ease of use of the TAM model. Larsen et al. (2009) combined TTF and post-acceptance model (PAM) to understand users’ motivation to continue the use of IS. They used TTF to predict the constructs’ perceived usefulness and utilization in the PAN model. Afshan & Sharif (2016) combined TTF, UTAUT, and ITM to investigate mobile banking acceptance in Pakistan. They used TTF to predict the behavioural intention construct of the UTAUT model. We thus find that most studies that used TTF had behavioural intention and adoption as the outcome. In this sense, the integration of TTF with ECM will provide some insight into mpayment research. 4.2.4. Trust Trust has been conceptualized in several ways according to the context in which it is applied (Gefen, Karahanna, & Straub, 2003; Oliveira, Alhinho, Rita, & Dhillon, 2017). According to early literature, in a general view trust reflects the ability of IS to fulfil the tasks correctly, i.e., the IS provider keeps their promise and does not deceive users, and the benefits of the IS need to be perceived by the users and the providers (Zhou, 2014). In m-payment, trust reflects users’ beliefs in the reliability of the transactions made through m-payment. If an m-payment provider ensures secure transactions, fulfils the tasks correctly, and does not deceive the users, it will be possible to improve the 71 Doctoral Programme in Information Management perception of reliability amongst the m-payment users (Chen & Li, 2017; Zhou, 2014). There are several research articles that combine trust with other theoretical models in IS research, such as Zhou & Li (2014) who investigate mobile social network services; Zhou (2013) to understand m-payment continuance intention; Gao et al. (2015) to perceive consumers’ mobile purchase continuance intention. In our research we use the overall trust and trust dimension. Benevolence, competence, and integrity comprise the dimension of trust (Oliveira et al., 2017). Competence reflects the ability of an IS provider to enforce their promises to users. Integrity reflects the IS provider’s capacity to act consistently, reliably, and honestly while keeping its promises. Benevolence reflects on the probability of an IS provider to maintain users’ interests and to show sincere concern with the well-being of the users (Chen & Dhillon, 2003; Oliveira et al., 2017; Palvia, 2009); 4.3. Research model The purpose of our study is to understand the continuance intention, which we use as the basis of the ECM model (A Bhattacherjee, 2001). Considering ECM an axiomatic theory, which is acceptable and truly self-evident (J. K. Lee et al., 2021), in the current study we adopted for our model two constructs, satisfaction and continuance intention (A Bhattacherjee, 2001), considering that the dependent construct of our study is continuance intention. Following the parsimony approach and reducing model complexity to make it easier to grasp, the constructs perceived usefulness and confirmation were not added. The addition of other models and variables offers a better understanding of continuance intention to use m-payment. We therefore join the TTF model, which directly influences use and individual performance, with trust, which is a direct determinant of use and continuance intention. The theoretical model (Fig. 4.1) is designed to examine the continuance intention to use m-payment in the African context. The model asserts that: 1. TTF determines the use of m-payment and the perceived individual performance: 72 Doctoral Programme in Information Management 2. the use of m-payment can have an indirect or direct influence on continuance intention, and a direct influence on individual performance; 3. individual performance determines the continuance intention directly; 4. the trust dimension can have an indirect or direct influence on continuance intention, and a direct influence on the use; 5. user satisfaction may moderate the impact of individual performance, use, and the trust dimension on m-payment continuance intention. The following section presents the proposed hypotheses. Figure 4.1 - Proposed research model. From a technical perspective, tasks are the activities carried out by users when they are using m-payment. However,, if the task characteristics of the m-payment are easy to use, appropriate and understandable (Wu & Chen, 2017), the users will be comfortable with their use and continue using it, on the other hand, users facing difficulties in using m-payment will prefer opting traditional payment method rather than m-payment (Rahi et al., 2020). Task characteristics are relevant constructs to influence positively TTF (Rahi et al., 2020; Tam & Oliveira, 2019). 73 Doctoral Programme in Information Management Technology characteristics are physical and logical tools (hardware and software), i.e., the look, feel, and speed of technology. An effective technology makes m-payment more attractive and useful for the users (Rahi et al., 2020). This factor can affect the long-term usage of the technology (Tam & Oliveira, 2016b, 2019). For m-payment, the technology characteristics play an important role, as users are carrying out monetary transactions, the speed of the operation and the response time are important factors. With the minimum technology characteristics, the users need to understand the actual working of tasks of m-payment (Tam & Oliveira, 2019; B. Wu & Chen, 2017). The TTF is the fit between task and technology characteristics (Goodhue & Thompson, 1995; Wu & Chen, 2017). This attribute means that when m-payment users perceive a match between task and technology (understand that the features and technological characteristics are suitable for carrying out transactions) it will be possible to improve the usage and the continued usage of m-payment (Tam & Oliveira, 2016b). Therefore, it is expected that the m-payment users will performs the tasks efficiently (Rahi et al., 2020). when the m-payment tasks are easy to use, fast, and provided anytime and anywhere, users will feel the usefulness of m-payment, thereby increasing the work performance of the individual. Thus, we hypothesise: H1a: TTF positively influences use. H1b: TTF positively influences individual performance. When users start to use any system, they start to perceive benefits. When using mpayment, users will perceive its benefits and the level of perceived individual performance will increase. Earlier studies report empirical support for the relationship between use and individual performance (Tam & Oliveira, 2016a, 2016b). Thus, the frequent use of m-payment to check balances, make transfers, pay for goods, etc. will influence the individual performance, and increase the intention to continue using mpayment. When users perceive effortless use, and when they start perceiving performance outcomes (Chang, Liu, & Chen, 2014), usage will become more frequent. When the use of m-payment services starts to become automatic and users use it more often, we expect that the continuance usage of m-payment will increase (Tam & Oliveira, 2016b). Thus, we hypothesize: 80 Doctoral Programme in Information Management 4.5.2. Structural model The results of the structural model were examined by the path coefficients that present the strength of the constructs’ relationships, variance inflation factor (VIF), t-statistic value, and variance explained (R2) to validate the hypotheses and constructs (see Fig. 4.2). The structural model examination used 5000 bootstrap resamples to estimate the paths’ significance (Henseler et al., 2009). We tested the VIF to assess the multicollinearity and all the constructs are below the threshold of 5, thus indicating the absence of multicollinearity (Hair et al., 2016). Figure 4.2 - Research model Regarding R2 (see detail in Fig. 4.2), the results of the PLS structural model explain 29.5% of the variation in use. The task technology fit, and overall trust are significant in explaining use. Thus, confirming H1a and H7a. The research model explains 45.2% of the variation in individual performance and are explained by task technology fit and use, confirming H1b and H2a. The research model explains 47.5% of overall trust variation and are explained by benevolence, competence, and integrity, confirming H4, H5, and H6. The research model explains 47.8% of the continuance intention, explained by use, individual performance, and overall trust, confirming H2b, H3, and H7b. Additionally, three moderating effects (H8a, H8b, and H8c) were examined. The 81 Doctoral Programme in Information Management findings showed that H8b and H8c were significant, and H8a was not statistically significant. However, the moderation effect of H8c is negative, meaning that greater user satisfaction will be weaker in the relationship between overall trust to continuance intention, thus confirming H8b and H8c. 4.5.3. Mediating role of use and individual performance The findings reveal that there are mediation effects on some constructs. Mediation effect (indirect effect) is presented by a third intervening variable between an independent and a dependent construct (Hair Jr. et al., 2016). We performed a mediation analysis and the results (Table 4.4) show that individual performance is a partial mediator between use and continuance intention. Also, use is a partial mediator between overall trust and continuance intention. Table 4.4 - Mediation analysis. Beta t-Test p-Value conclusion H9a: OT → U → CI 0.103 3.922 0.000 Partial mediation H9b: U → IP → CI 0.096 3.635 0.000 Partial mediation Notes: Use (U); individual performance (IP); overall trust (OT), and continuance intention (CI). 4.6. Discussion We developed and validated a conceptual model that explains the importance of the TTF model and overall trust toward continuance intention to use m-payment. The findings show that 13 of the 15 hypotheses were confirmed. Overall trust can influence use and continuance intention to use m-payment. Overall trust is supported by benevolence, competence, and integrity. This means that m-payment service providers should ensure the best interests of their end-users, service support should do its best to assist users, and end-users should feel supported and confident with the services. Additionally, the service provider must be honest with users, keeping its commitments with the end-users. When the m-payment service provider has enough expertise to support users, they will start trusting the provider and the technology, thereby motivating users to use m-payment, and boosting their intention to continue using mpayment (Oliveira et al., 2017; Yu et al., 2018; T. Zhou, 2011b). By improving users’ 82 Doctoral Programme in Information Management overall trust, it will be possible to improve use and the intention to continue using mpayment (Zhou, 2013, 2014). TTF positively impacts use and individual performance. The task and technology characteristics are fundamental to the use of m-payment, considering that users have different experiences. Users must realize that the features are objective, easy to use, with perceptible information, and that the characteristics of the technology are adequate to use the functionalities. For example, in Mozambique, service providers should ensure a greater fit of task and technology characteristics to enhance the usage of m-payment by users (Tam & Oliveira, 2016b). Considering the limitations of banking infrastructure, in terms of space and time, mpayment is a useful alternative, as it provides real-time services anytime and anywhere and is an attractive alternative for people who live far from banks (Yu et al., 2018). By improving the task and technology characteristics, it will be possible to improve the use and perceived individual performance of end-users, and consequently, improve the intention to continue using m-payment. Use positively affects individual performance and continuance intention to use m-payment. However, when the users sense trustworthiness and m-payment providers deliver adequate services with good characteristics to end-users, they will feel satisfied and motivated to use m-payment, consequently perceiving the performance of m-payment and being satisfied to continue using it (Fan, Shao, Li, & Huang, 2018; Liébana-Cabanillas et al., 2018). Individual performance positively influences continuance intention to use m-payment. When endusers understand that m-payment is useful for their daily financial activities, helps to accomplish their tasks easily, and enables them to do tasks more quickly, users perceive individual performance, and continue using m-payment. The m-payment provider should ensure good m-payment characteristics that are easy to use, easy to interpret, fast, and available anywhere and anytime. Doing so will allow the users to perceive the benefits, guaranteeing perceived individual performance (Larsen et al., 2009; Tam & Oliveira, 2016b). By improving individual performance, it will be possible to improve the intention to continue using m-payment. Table 4.5 - Results of the hypotheses. Hypotheses Independent construct → Dependent construct Findings (β) P-value Support H1a Task technology fit → Individual performance 0.500 0.000 Yes H1b Task technology fit → Use 0.342 0.000 Yes 83 Doctoral Programme in Information Management H2a Use → Individual performance 0.275 0.000 Yes H2b Use → Continuance intention 0.328 0.000 Yes H3 Individual performance → Continuance intention 0.350 0.000 Yes H4 Benevolence → Overall trust 0.333 0.000 Yes H5 Competence → Overall trust 0.246 0.000 Yes H6 Integrity → Overall trust 0.247 0.000 Yes H7a Overall trust → Use 0.315 0.000 Yes H7b Overall trust → Continuance intention 0.099 0.071 Yes H8a Individual performance * Satisfaction → Continuance intention 0.017 0.778 No H8b Use * Satisfaction → Continuance intention 0.099 0.033 Yes H8c Overall trust * Satisfaction → Continuance intention -0.126 0.025 No Additionally, satisfaction moderates the relationships among use and overall trust of mpayment to explain continuance intention (Susanto et al., 2016; Yu et al., 2018). When the use of m-payment is moderated by the existence of satisfaction, it is observed that the impact will be high to explain continuance intention to use m-payment. In contrast, the moderating effect of satisfaction on overall trust to explain continuance intention will be weaker. In this sense, if m-payment users have a high level of satisfaction, use will gain strength, and overall trust will lose strength in explaining m-payment continuance intention. The findings also reveal that individual performance is a partial mediator between use and continuance intention, and use is a partial mediator between overall trust and continuance intention. When users have confidence of m-payment, in the services, and value the reliability properties, they increase their intention to continue using m-payment. With use mediation and frequent use of m-payment, trust improves, and consequently intention to continue using m-payment increases. This means that when the users use m-payment in their daily lives to buy products or services and transfer or withdraw money, the intention to continue using it increases. During mpayment use, perceived individual performance such as accomplishing tasks quickly, easily, and perceiving the usefulness of m-payment in everyday life, enhances the positive impact of continuance intention to use m-payment. Thus, m-payment providers should ensure fast task completion, ease of use, and provide services that are useful for the user. We plotted the moderations of satisfaction to better understand its behaviour (Fig. 4.3) (Aiken et al., 1991). Figure 4.3 illustrates that use of m-payment has a more significant 84 Doctoral Programme in Information Management impact on m-payment continuance intention when user satisfaction is higher. Thus, with a higher level of user satisfaction, the use of m-payment will increase the intention to continue using m-payment. Additionally, overall trust has a low significant impact on m-payment continuance intention when user satisfaction is higher. Therefore, the importance of overall trust for m-payment continuance intention is important when user satisfaction is low. Figure 4.3 - Moderation effect of user satisfaction 4.6.1. Theoretical implications The current study investigates the continuance intention to use m-payment. Therefore, we integrated TTF, overall trust and ECM. Our results indicate that both TTF and overall trust are important and should be considered when evaluating the continuance intention to use m-payment or similar systems. This means that users' perceptions associated with the level of fit between features and technology, as well as users' confidence in using m-payment may ultimately lead to the intention to continue using m-payment. The study joins the TTF, overall trust, and ECM models to evaluate mpayment, which has not been reported in previous literature. TTF and overall trust were combined to explain continuance intention of the ECM, which is, nevertheless, a wellknown model and is one of the most popular and widely accepted models (A Bhattacherjee, 2001). The findings suggest that the integration of TTF and overall trust present predictive power to explain continuance intention. Individual performance, use, and overall trust explain 47.8% of the variation in continuance intention. This result indicates that our model performs well compared to previous studies such as the original ECM (A Bhattacherjee, 2001), which explained R2=41%, Idemudia et al. 85 Doctoral Programme in Information Management (2018), which explained R2=46%, Albashrawi and Motiwalla (2019), which explained R2=45.9%, and Gong et al. (2020), which explained R2=41.6%. Thus, indicating that TTF and overall trust are important antecedents for m-payment continuance intention. Second, the model was applied in the African context for m-payment, addressing the concept of continuance intention (Humbani & Wiese, 2019). To the best of our knowledge, very few studies have addressed continuance intention in this context. Therefore, with the proposed model, researchers in the IS field can adapt it to suit other situations in the future. Third, the findings suggest that individual performance and use are the strongest predictors of continuance intention in the context of m-payment. Nevertheless, the results show that the constructs of overall trust and satisfaction must be taken into consideration when addressing continuance intention (A. Bhattacherjee & Lin, 2014; T. Zhou, 2014b). In the Bhattacherjee (2001) model satisfaction is the strongest predictor of continuance intention. In our model satisfaction was explored as a moderator of use, individual performance, and overall trust in continuance intention. The results show that satisfaction moderates the relationship between use and overall trust on continuance intention. Interestingly, the moderation effect in the relationship of overall trust on continuance intention is negative, suggesting that when the level of satisfaction is high, trust is not an important factor influencing continuance intention, showing that user trust is important only when the level of satisfaction is low. The results also suggest that the constructs of individual performance and use are partial mediators of use and continuance intention, and overall trust and continuance intention, respectively. This study demonstrates that the proposed model provides support for the importance of the added constructs, such as the trust dimension to explain continuance intention. The study shows that TTF is an important predictor to use of m-payment and perceived individual performance (Oliveira, Faria, et al., 2014; Tam & Oliveira, 2016b). Use is an important predictor of perceived individual performance and mpayment continuance intention. Overall trust, use, and perceived individual performance are important predictors of continuance intention. 86 Doctoral Programme in Information Management 4.6.2. Practical implications Our study has several practical implications for m-payment decision-makers and providers. Our results suggest that m-payment providers seeking long-term usage should focus on real-time accessibility, real-time services, and services that are quick and secure in order to enhance the task and technology fit of m-payment (Ouyang et al., 2017). This finding is very important to decisionmakers and providers because when they provide services with a better fit between task and technology (e.g., providing tasks that are easy to use, enhancing system speed, reducing system downtime, etc.), it will affect the use of m-payment, increase the perceived individual performance, and consequently enhance m-payment continuance intention. In this sense, if m-payment providers want their active customers (users) to continue using mpayment, they should provide adequate services to them. This study implies that benevolence, competence, and integrity have a significant impact on overall trust (Oliveira et al., 2017; Tam et al., 2019). This, in turn, suggests that m-payment providers should handle m-payment transactions, be truthful to users, act genuinely with users, keep their commitments, and do their best to help users, especially when it involves fraud, or pending transactions. Doing these, it will be possible to increase users’ trust in m-payment. Perceived benevolence occurs when the user believes that the m-payment provider acts in their best interest, and if needed the provider will do their best to help. Perceived competence is when the user believes that the m-payment provider has the ability to handle m-payment transactions. Perceived integrity occurs when the user believes that the m-payment provider acts sincerely, is honest, and keeps to their commitments. Therefore, in order to improve the reliability for users, the m-payment provider should be honest and user-oriented to create a good image, because users need to believe that the m-payment provider will always fulfil their promises. This might encourage users to use and continue using m-payment (Oliveira et al., 2017; Palvia, 2009). However, to enhance the continuance intention to use m-payment, the provider should ensure trustworthiness and provide quick and easy to use tasks that are easy to understand in order to ensure perceived individual performance of users. Considering 87 Doctoral Programme in Information Management the moderating effects of satisfaction, the m-payment provider should ensure a high level of satisfaction to improve the use of m-payment. When the users are satisfied, they will use and also invite others to use m-payment. Based on these findings, it is recommended that the m-payment provider base their action plans on the determinants that influence m-payment users, such as task-technology fit, trust, perceived individual performance, and the use of m-payment. 4.6.3. Limitations and future research Some limitations exist in our study. The data were collected in Mozambique, and to generalize the applicability of the study, it is suggested that future studies could be conducted in other African countries. The sample represents a highly educated population because we collected the data at universities, but most m-payment users are from the rural areas of Mozambique. Future studies may test our model in a different part of the country and/or in another African country. Considering that gender equality is an interesting and important topic in the African context (Humbani and Wiese, 2018), future research may examine the differences between genders. This study is related to a single type of technology (m-payment); a comparison with another technology (e.g., m-banking) might reveal other insights and enhance generalization. Considering that cultural factors play an important role in an African context, future research might include some cultural factors, such as uncertainty avoidance or individualism. 4.7. Conclusions The rise of m-payment in Africa has brought many opportunities to people, and banks are changing the way that they provide services to local communities (Batista & Vicente, 2018; Humbani & Wiese, 2018). Our study empirically assesses TTF and the dimension of trust to create the intention to continue using m-payment amongst users. This study contributes to the literature by providing a theoretical model to explain continuance intention to use m-payment. It provides a baseline to decision-makers and providers of how technological factors and trust are important to ensure long-term usage of m-payment. The results demonstrate that the most important factors to explain m-payment continuance intention are individual performance, use, and overall trust. 88 Doctoral Programme in Information Management Individual performance and use play important roles as partial mediators between use – continuance intention and overall trust – continuance intention. Our results show that satisfaction has significant importance as a moderator between use and overall trust on continuance intention. 89 Doctoral Programme in Information Management 96 Doctoral Programme in Information Management Figure 5.1 - Proposed model The confirmation refers to the expectations that the users have in using the IS. The users make their evaluation when comparing their initial benefits with the expected benefits. The ECM shows that the positive effects of confirmation will impact the satisfaction and perceived usefulness (Alraimi et al., 2015; Oghuma et al., 2016). When the initial expectation of users is confirmed, it affects the level of the user satisfaction and perceived benefits of IS (Susanto et al., 2016). For m-payment, the user who confirmed the expectations can realize the benefits and influence the satisfaction. Therefore, we hypothesize: H1. Confirmation positively impacts the m-payment perceived usefulness. H2. Confirmation positively impacts the m-payment satisfaction. The determinant that affects the users to consider that IS enhances their effectiveness, performance, or productivity is perceived usefulness (C. M. Chiu & Wang, 2008; Davis, 1989). This means, when users perceive the benefits of the IS, the long-term usage is reinforced (Lee, 2010; Rezvani et al., 2017). ECM postulates that when the expected benefits are confirmed, the user realizes the advantage, which consequently influences positively their satisfaction and long-term usage. In our study, when the m- 97 Doctoral Programme in Information Management payment user feels that using m-payment is useful and enhances his or her performance, he or she will be more satisfied and will continue using it (Cho, 2016; Joo et al., 2018; Shin et al., 2017). We hypothesize: H3. Perceived usefulness positively impacts the m-payment satisfaction. H4. Perceived usefulness positively impacts the m-payment continuance intention. Satisfaction is the extent to which a user acquires a positive feeling by using m-payment resulting from the usage experiences and performance outcomes (A. Bhattacherjee & Lin, 2014). When the IS leaves the user satisfied, the long-term relationships become stronger (Yu et al., 2018). The ECM postulates that user satisfaction is the central reason for the continuance intention of IS (A Bhattacherjee, 2001). Bhattacherjee (2001) validated empirically that the relationship satisfaction on continuance intention was the strongest. Other research also demonstrate that satisfaction of the user strongly influence intention to continue to use IS (Cho, 2016; Joo et al., 2018; Shin et al., 2017). In our study, if the m-payment user feels satisfied, the long-term usage will be guaranteed. Therefore, we hypothesize: H5. Satisfaction positively impacts the m-payment continuance intention. 5.3.1. Moderation role of uncertainty avoidance Confirmation refers to the user’s evaluations toward a product, service, or technology, whether it is positive or negative. The positive confirmation is when the user reaches the initial expectations, while the negative confirmation is when the user does not reach the initial expectations (Alraimi et al., 2015; Oghuma et al., 2016). Users make their assessments when comparing their initial expectations with the efficiency of the product, service, or technology. The confirmation of expectation is when m-payment services enhance the user’s perceived usefulness and satisfaction of the service (Susanto et al., 2016). Given the cultural aspect of the users, they will have attitudes that vary according to the context. 98 Doctoral Programme in Information Management Uncertainty avoidance is the attitude that the user takes to avoid ambiguous or unknown situations (Hofstede, 1984). A low level of uncertainty avoidance means that people are not averse to taking risks. Thus, there is a greater degree of acceptance for new technologies. The low-level of uncertainty avoidance of users of a given cultural context is a factor that supports the relationship between the user satisfaction and confirmation. Uncertainty avoidance also influences the relationship among perceived usefulness and confirmation. The influence of the uncertainty avoidance on these relationships increases user’s satisfaction and perceived usefulness. We hypothesize: H6a. Uncertainty avoidance positively moderates the impact of the confirmation on perceived usefulness of m-payment. H6b. Uncertainty avoidance positively moderates the impact of the confirmation on satisfaction of m-payment. The factor that may help an individual to start to understand the advantages in terms of utilization of the IS is perceived usefulness (Davis, 1989). This means, when the users perceive the improvement of using the services and system, the long-term relationship is reinforced (Lee, 2010; Rezvani et al., 2017). However, for m-payment, perceived usefulness is important because it enables the frequent use of m-payment. Given the cultural context of the users, the uncertainty avoidance level can be low or high. Low level is an indicator that users will use m-payment services with little hesitation, while high-level of uncertainty avoidance is an indicator that users will use m-payment services with much hesitation. The user uncertainty avoidance low level of a given cultural context is a factor that favours the relationship among the satisfaction and perceived usefulness. The uncertainty avoidance also influences the relationship among perceived usefulness and the intention to continue to use m-payment. We hypothesize: H6c. Uncertainty avoidance has a positive impact on the moderation of perceived usefulness on satisfaction of m-payment. H6d. Uncertainty avoidance positively impacts the moderation of perceived usefulness on m-payment continuance intention. 99 Doctoral Programme in Information Management Satisfaction is the degree to which an individual is comfortable using m-payment services due to usage experiences and performance outcomes, meaning that satisfaction starts to become stronger after the users adopt the service or system (A. Bhattacherjee & Lin, 2014). When users are satisfied with the service, long-term relationships become stronger (Yu et al., 2018). The level of uncertainty avoidance may influence the increase or decrease of the impact on the relationship between satisfaction and the mpayment continuance intention. Given the cultural context of the users, the low level of uncertainty avoidance will positively influence the relationship between satisfaction and the m-payment continuance intention. We hypothesize: H6e. Uncertainty avoidance positively impacts the moderation of satisfaction and the m-payment continuance intention. 5.4. Research methods The current research employs a mixed-methods approach (Soffer & Hadar, 2007; Venkatesh et al., 2013). First, was applied the quantitative method, in which the main method of data collection was an online survey (Alraimi et al., 2015; Tam & Oliveira, 2016b). A questionnaire was constructed for the survey using variables and items from the research (Appendix A). The measurement items of the model were adopted from published studies and for uncertain avoidance we adopted from Srite and Karahanna (2006). Items for continuance intention, satisfaction, perceived usefulness, and confirmation were adapted from Bhattacherjee (2001). Secondly, we applied a qualitative method to triangulate and obtain additional impressions regarding the findings, and we thus employed field interviews (Venkatesh et al., 2013). 5.4.1. Data For the quantitative method, a seven-point scale was applied to assess the items, ranging from 1 (totally disagree) to 7 (totally agree). The questionnaire was managed in English and reviewed for content validity by a professional. A professional translator translated 100 Doctoral Programme in Information Management to the Portuguese language, taking into consideration that the survey was administered in Mozambique. The questionnaire was reverse translated to the original language (Brislin, 1970). To validate the instruments, we conducted a pilot test on a group of 30 students (these data were excluded from the final analysis). To elevate the response rate, various strategies were used. First, we applied the “key informant” procedure to collect data (Oliveira, Thomas, et al., 2014; Pinsonneault & Kraemer, 1993), which helped in the identification of qualified respondents. To boost responses to the survey, a follow-up message was sent two months after the initial contact. A total of 384 usable responses were obtained from which 272 were part of the first respondents and 112 were part of the respondents reminded by the follow-up email. Comparing the first and last respondent groups using Kolmogorov–Smirnov (K–S), the tests indicated a lack of non-response bias (Ryans, 1974). Two tests were made to detect the common method bias. Harman’s one-factor test (Podsakoff et al., 2003). The first construct explains 36.7% of variance, it acknowledges that any of the constructs individually account for greater variance. Second, using a marker variable procedure (Lindell & Whitney, 2001), adding a theoretically irrelevant marker variable in the study’s model, obtaining 0.040 (4.0%) as the maximum shared variance with other variables; the value can be considered as low (Johnson et al., 2011). We therefore found no significant common method bias. The data were collected between July 2018 and January 2019. Statistics show that 64% of the respondents were professional workers, of which 59% were men, and that 39% had used m-payment one to four times during the preceding three months (see Table 5.1). For the qualitative method we employed field interviews to explain the motivations of users regarding the intention to continue to use m-payment (Soffer and Hadar, 2007). To collect different perspectives of m-payment, we interviewed seven people, selected randomly, the first and second interviewees (I1 and I2) were information technology (IT) students, I3 a bank officer, I4 an IT technician, I5 an entrepreneur, and I6 and I7 were university professors. M-payment users are identified as individuals who use mpayment services (e.g., M-Pesa or Mkesh in Mozambique). The interviews were conducted using the items of the theoretical model proposed in the study. 101 Doctoral Programme in Information Management Table 5.1 - Sample characteristics Age < 25 129 34% 25 - 30 122 32% 31 - 40 85 22% 41 - 50 39 10% > 50 9 2% Gender Female 158 41% Male 226 59% Education High school or below 91 24% Bachelor’s degree 179 47% Master's degree or higher 114 30% Employment Students 99 26% Professional workers 244 64% Retired 1 0% Unemployed 40 10% Marital status Single 187 49% Married 86 22% Divorced 23 6% Widowed 10 3% Marriage in fact (cohabitation) 75 20% Do not know answers 3 1% M-payment usage frequency (time / 3 months) 1 - 4 149 39% 5 - 10 98 26% > 10 137 36% 5.4.2. Data analysis and results To test and assess the research hypotheses of the model, we applied partial least squares - structural equation modelling (PLS-SEM), Smart PLS 3 (Ringle et al., 2015). Earlier research has recognized the potential of PLS-SEM for theory development (Alraimi et al., 2015; Côrte-Real et al., 2020; Tam & Oliveira, 2016b). Additionally, the data do not have normal distribution, the research model is complex, and has not been tested in previous research. Therefore, PLS-SEM is appropriate for the current research. 5.4.3. Measurement model The measurement model assessment was conducted by applying (1) indicator reliability (considering good loading greater than 0.70, and thus excluding CI4), (2) construct reliability (using composite reliability (CR) indicator, good CR greater than 0.70), (3) 102 Doctoral Programme in Information Management convergent validity (using average variance extracted (AVE), good AVE greater than 0.50) (Fornell & Larcker, 1981; Hair Jr. et al., 2016; Henseler et al., 2009), and (4) discriminant validity. The findings are shown in Tables 5.2 and 5.3. Table 5.2 - Construct Reliability and Validity Constructs AVE Composite Reliability Cronbach's Alpha Item Loadings t-value Confirmation 0.666 0.857 0.749 C1 0.820 35.782 C2 0.839 35.733 C3 0.787 27.602 Perceived usefulness 0.615 0.865 0.791 PU1 0.767 25.937 PU2 0.805 31.442 PU3 0.810 37.051 PU4 0.755 22.124 Satisfaction 0.721 0.886 0.807 S1 0.859 55,392 S2 0.858 52.386 S3 0.832 37,340 Continuance intention 0.567 0.834 0.739 CI1 0.829 44.778 CI2 0.815 30.272 CI3 0.831 45.300 Uncertainty avoidance 0.631 0.873 0.806 UA1 0.788 33.868 UA2 0.809 34.681 UA3 0.807 28,310 UA4 0.774 24,918 Table 5.3 - Fornell-Larcker Criterion Mean STDEV C PU S CI UA Confirmation 4.449 1.198 0.816 Perceived usefulness 4.597 1.175 0.631 0.784 Satisfaction 4.610 1.250 0.676 0.600 0.849 Continuance intention 4.635 1.144 0.469 0.489 0.462 0.753 Uncertainty avoidance 4.923 1.186 0.346 0.357 0.281 0.459 0.794 All constructs met the above-described standards, thereby ensuring convergence. This shows that the factors can be applied to test the theoretical model. The Fornell-Larcker test was used to assess (4) discriminant validity. The AVE square root of each factor should be greater than the correlation among the factors (Fornell and Larcker, 1981) (see Table 5.3). Further, we examined cross-loadings criteria (Appendix B), and the heterotrait-monotrait ratio of correlations (HTMT) (Appendix C). The items show higher loading on their corresponding constructs than the cross-loadings, thereby ensuring discriminant validity (Chinn, 1998; Götz et al., 2010). All of the HTMT values are below 0.90, thus concluded for discriminant validity (Henseler et al., 2015). The 103 Doctoral Programme in Information Management results indicate that the factors are statistically different and can be assessed in the structural model. The measurement model findings show a good indicator reliability, construct reliability, convergence validity, and discriminant validity. 5.4.4. Structural model The structural model was examined after the confirmation of the measurement model, using the path coefficients, variance inflation factor (VIF), t-statistic value, and variance explained (R2) to test the hypotheses and the constructs (see Figure 5.2). The structural model used 5,000 bootstrap resamples to evaluate the paths’ significance (Henseler et al., 2009). VIF was used to assess the multicollinearity, all constructs met the criteria with values below 5, and thus it can be concluded that there is an absence of multicollinearity (Hair Jr. et al., 2016). Figure 5.2 - Research model (*=p<0.10; **=p<0.05; ***=p<0.01). The structural model explains 43.0% of the variation in perceived usefulness. The confirmation ( " # = 0.556, p < 0.01) is statistically significant in explaining perceived usefulness, supporting hypothesis H1. Moreover, the research model explains 52.4% of the variation in satisfaction. Confirmation ( " # = 0.479, p < 0.01) and perceived 104 Doctoral Programme in Information Management usefulness ( " # = 0.261, p < 0.01) are statistically significant in explaining satisfaction, supporting hypotheses H2 and H3. The research model explains 38.2% of variation in continuance intention, explained by perceived usefulness ( " # = 0.201, p < 0.01) and satisfaction ( " # = 0.224, p < 0.01), supporting hypotheses H4 and H5. Additionally, five moderating models were examined (H6a, H6b, H6c, H6d, and H6e). The results show that H6a ( " # = 0.081, p < 0.05), H6b ( " # = 0.101, p < 0.10), and H6d ( " # = 0.127, p < 0.05) were statistically significant, thus supporting hypotheses H6a, H6b, and H6d. The hypotheses H6c ( " # = 0.007, p < 0.10) and H6e ( " # = -0.025, p < 0.10) were not statistically significant. 5.5. Discussion This study assesses the impact of cultural dimension uncertainty avoidance on the ECM model for m-payment. As far as we know it is the first empirical research that examines ECM taking into consideration the moderation effect of uncertainty avoidance. By applying a mixed-methods approach we were able to explore the qualitative view of the antecedents of the intention to continue to use m-payment. Table 5.4 shows a summary of the findings of the hypothesis’s conclusions. As expected, all of the model relationships were supported (A Bhattacherjee, 2001). This finding is in line with those reported in other studies (X. Lin et al., 2017; Susanto et al., 2016). Furthermore, our findings indicate that confirmation of expectation positively impacts the satisfaction and perceived usefulness of m-payment. This means that the experiences using mpayment were positive and user expectations were confirmed (Cheng, 2014; Tam et al., 2020). In a qualitative view, the interviewees I1, I2, and I5 (IT students and entrepreneur) highlighted the perception of usefulness and the feeling of satisfaction with m-payment, stating that they can easily pay for school and food expenses. The I5 (entrepreneur) highlighted the ease of selling products, because most customers chose to pay with m-payment because of the facilities, and they did not use the traditional method via cash, mainly because of the COVID-19. They also reported the facilities to carry out transactions, anywhere and anytime, offering greater availability, fast and safe, simplifying their daily lives, especially for those who live far from urban centres. 105 Doctoral Programme in Information Management The perceived usefulness positively impacts the satisfaction and continuance intention. Meaning that when users perceive performance, effectiveness, or benefits on mpayment, satisfaction will be confirmed, consequently influencing continuance intention. All the interviewees remarked extensively about the usefulness of mpayment, in different aspects; for the IT students (I1 and I2) it is useful for paying fees and purchasing food; for bankers it is useful because it reduces the number of people needing to use bank services; and also, for personal expenses, goods, and services such as gas, water, and electricity. Moreover, the interviewees argued that m-payment has great utility, especially nowadays with the COVID-19 pandemic, facilitating transactions from home or anywhere and anytime, reducing distances, and saving time – thus, allowing to allocate time for other activities and helping to reduce the risk of contamination by the coronavirus. The results indicate that satisfaction positively influences continuance intention. This means that when the user is happy and delighted, they will continue to use m-payment. The interviewees I1, I2 and I5 (IT students and entrepreneur) greatly emphasized their satisfaction with m-payment and argued that it is a very important factor for the intention to continue to use m-payment. They mentioned ease of use and the ability to use it anywhere and anytime, especially at this time of the COVID-19 pandemic. They indicated their satisfaction because they can pay basic expenses such as water, energy, and TV from home, thereby reducing travel and the number of coronavirus infections. The m-payment makes user’s lives much easier. It is widely used at the national level. The pandemic accelerated use, mainly because of government-imposed containment rules and because the users should avoid touching objects such as ATM, POS, doorknobs, etc. Table 5.4 - Hypotheses results Hypotheses Independent Construct → Dependent construct Findings (β) P value Conclusion H1 Confirmation → Perceived usefulness 0.556 < 0.01 Supported H2 Confirmation → Satisfaction 0.479 < 0.01 Supported H3 Perceived usefulness → Satisfaction 0.261 < 0.01 Supported H4 Perceived usefulness → Continuance intention 0.201 < 0.01 Supported H5 Satisfaction → Continuance intention 0.224 < 0.01 Supported 112 Doctoral Programme in Information Management Chapter 6 – Conclusions 6.1. Summary of results The main objective of this paper is to determine the main determinants of the continuance intention to use m-payment. We conducted four quantitative studies (Chapter 2 to Chapter 5), one literature review and three empirical studies that analyse the effects of different factors and models on the intention to continue using mpayment. Table 6.1 summarizes the statistical results of the relationships of the different models analysed in the studies. Table 6.1 - Relationships analyses in all the studies Independent Construct Dependent construct Chapter 2 Chapter 3 Chapter 4 Chapter 5 Continuance intention Continuance behaviour 0.375 Attitude Continuance intention 0.441 Flow Continuance intention 0.358 Subjective norms Continuance intention 0.179 Hedonic value Continuance intention 0.437 Utilitarian value Continuance intention 0.242 Affective commitment Continuance intention 0.556 Trust Continuance intention 0.239 Perceived enjoyment Continuance intention 0.187 Performance Continuance intention 0.241 Habit Continuance intention 0.255 Perceived behaviour control Continuance intention 0.296 Use Continuance intention 0.272 0.328 Satisfaction Continuance intention 0.416 0.162 0.224 Perceived usefulness Continuance intention 0.285 0.016 (ns) 0.201 Individual performance Continuance intention 0.310 0.350 Overall trust Continuance intention 0.099 Service quality Satisfaction 0.067 (ns) System quality Satisfaction 0.279 0.065 (ns) Confirmation Satisfaction 0.431 0.389 0.479 Use Satisfaction 0.240 Perceived usefulness Satisfaction 0.248 0.136 0.261 Information quality Satisfaction 0.248 0.152 Disconfirmation Satisfaction 0.576 Perceived ease of use Satisfaction 0.188 113 Doctoral Programme in Information Management Perceived enjoyment Satisfaction 0.251 Confirmation Perceived usefulness 0.504 0.566 0.556 Disconfirmation Perceived usefulness 0.174 Perceived ease of use Perceived usefulness 0.327 System quality Confirmation 0.116 (ns) Service quality Confirmation 0.275 Information quality Confirmation 0.359 Use Individual performance 0.416 0.275 Satisfaction Individual performance 0.146 Task technology fit Individual performance 0.500 System quality Use 0.088 (ns) Service quality Use 0.225 Information quality Use 0.400 Task technology fit Use 0.342 Overall trust Use 0.315 Benevolence Overall trust 0.333 Competence Overall trust 0.246 Integrity Overall trust 0.247 Relational capital Affective commitment 0.258 Utilitarian value Affective commitment 0.112 Hedonic value Affective commitment 0.339 Satisfaction Attitude 0.481 Perceived usefulness Attitude 0.408 Confirmation Perceived enjoyment 0.622 Confirmation Perceived ease of use 0.458 Note: ns (not statistically significant) The results of the literature review show that the most used technologies were elearning and social network services. The most used theoretical models to study the intention to continue to use were ECM, TAM, and ECT. Additionally, few studies used a single theory, most studies integrated more than one theory. A theoretical model was presented based on the significant constructs from meta-analysis and best predictor from weight analysis. Of the 600 relationships collected from 115 studies, 60 relationships were analysed three or more times. Of these relationships, 31 were classified as “best predictors” and 24 were classified as “promising predictors”. The most studied regions were East Asia, North America, Europe, Middle East, South America, and Southwest Asia. No studies were found for the African region. 114 Doctoral Programme in Information Management In the first empirical study, the intention to continue using m-payment in an African context was analysed, combining ECM with D&M ISSM. The results show that the use of m-payment, satisfaction, and perceived individual performance are the most important factors to explain the continuance intention to use m-payment. Information quality and service quality positively impact the use and confirmation of the expectations. Information quality, use, and confirmation of expectations positively impact user satisfaction. In the second empirical study, intention to continue using m-payment was analysed by integrating TTF, overall trust, and ECM. The results show that the most important factors explaining the intention to continue using m-payment are individual performance, use, and overall trust. Individual performance and use play important roles as partial mediators between use-continuance intention and overall trustcontinuance intention. Our results show that satisfaction has significant importance as a moderator between use and overall trust on continuance intention. In the last empirical study, the impact of culture on the intention to continue using mpayment was analysed using a mixed-methods approach based on quantitative data and field interviews. The results show that the cultural factor uncertainty avoidance moderates the relationships between confirmation on satisfaction and confirmation on perceived usefulness, and perceived usefulness on continuance intention. In a qualitative view, most of the interviewees highlighted the importance of m-payment in their daily lives, especially for those who live far from urban centres. Thus, it was found that the culture plays a significant role in ensuring usefulness, satisfaction, and the continuance intention to use m-payment. 6.2. Contributions 6.2.1. Implications for theory This dissertation provides several contributions for research. The quantitative approach of the literature review contributes to research by providing a more concise, clearer, and extensive image of the constructs evaluated in previous studies in various subject areas on continuance intention to use IS from years 2001 to 2017, serving as the basis 115 Doctoral Programme in Information Management for future research and contributing for researchers to accurately select the constructs to be included in research models to assess m-payment continuance intention. Moreover, the results show that there is a wide variety of constructs coming from different theories and contexts that can significantly influence continuance intention, thus demonstrating that different theories and self-constructs can be integrated into the ISCI context. In Chapter 3 we proposed the joining of D&M ISSM and the ECM model, with the aim of identifying antecedents that focus on satisfaction, individual performance, and continuance intention. The findings provide support for the importance of the added constructs from DeLone and Mclean (2003) in users’ continuance intention to use mpayment. As another theoretical implication the model validates IS continuance intention theory for the case of m-payment use in Mozambique, unlike previous studies that focused on developed economies (X. Chen & Li, 2017; Fan et al., 2018a). In Chapter 4 we advanced the body of knowledge of m-payment by proposing the investigation of continuance intention to use m-payment integrating ECM, TTF, and overall trust. The results indicate that both TTF and overall trust are important and should be considered when evaluating the continuance intention to use m-payment. To the best of our knowledge very few studies have addressed continuance intention in the African context. Thus, with the proposed model researchers in the IS field can adapt it to suit other situations in the future. In Chapter 5 we presented the effects of the uncertainty avoidance cultural dimension in the ECM model using the m-payment case, applying mixed-methods, and were able to offer a holistic view of the antecedents of the continuance intention. The current research enriches the body of literature on ECM by extending the scope of IS continuance intention to the m-payment context and reveals the moderating role of uncertainty avoidance in predicting continuance intention. Considering that Mozambique is still consolidating the adoption phase of m-payment (Batista & Vicente, 2014, 2018), the cultural factor plays a significant role in the moderation of the continuous use of m-payment. 116 Doctoral Programme in Information Management 6.2.2. Implications for practice The results of this dissertation have valuable practical implications for managers and decision makers to ensure the users’ retention and long-term usage of m-payment. First, according to the findings of our research it is crucial to recognize the best predictors of continuance intention to use an IS, for better design and implementation of the IS (A. Bhattacherjee, 2001; Shao, 2018; Yu et al., 2018). The continuance intention to use an IS has been studied in different countries with different cultures. Therefore, managers should have different managing strategies to ensure the satisfaction of the users and long-term usage of the IS (L. Zhang et al., 2012). Managers and decision makers should provide all the necessary information, such as user guides, advertisements, and flyers that explain the services and functionalities of the IS, to accelerate the understanding of the services. Second, the m-payment providers should ensure that the information available is correct, up-to-date, and useful to the user. In addition, m-payment providers should help the users whenever they need help, should ensure that users have a good experience with m-payment, should provide services that exceed users’ expectations, and ensure that the users can use m-payment easily. This suggests that the m-payment providers should constantly improve the m-payment in features related to safety, ease of use, and information, to provide a well-structured system that is easy to navigate and has useful information. Third, to promote long-term usage of m-payment, providers should focus on real-time accessibility, real-time services, and services that are quick and secure in order to enhance the task and technology fit of m-payment (Ouyang et al., 2017). In other words, if the provider’s goals are to influence the active customers (users) to continue using m-payment, they should provide adequate services to them. Moreover, the m-payment providers should handle m-payment transactions, be truthful to users, act genuinely with users, keep their commitments, and do their best to help users, especially in situations involving fraud, or pending transactions. Fourth, the findings regarding the impact of culture on continuance intention to use mpayment indicate that managers should provide users with guidelines including all the instructions, procedures, and regulations to facilitate and help the users to operate with 117 Doctoral Programme in Information Management m-payment. With the support of instruction manuals, users will more easily and frequently use m-payment, and over time will come to perceive its benefits and the increase of efficiency it provides. Understanding culture could be important in the development and management of solutions for m-payment. For example, with the high level of uncertainty avoidance, managers could focus on mitigating uncertainty and risk, and ensuring the comprehension of the service for users in order to positively influence continuous use. 6.3. Limitations and future research The present dissertation contains several limitations. Regarding the literature review, we excluded certain studies because of the unavailability of their quantitative data, or because they were qualitative. Including these studies could generate relevant information in terms of the significance of the constructs. We used meta-essentials, a meta-analysis tool. This tool has limitations, and considering that meta-essentials is not able to perform more advanced analyses with linear models or structured equation modeling (Rhee et al., 2015), future researchers should consider a more advanced tool to provide more insights and a different approach to research. Regarding the three empirical studies presented in Chapters 2, 3, 4, and 5, the data were collected in universities, thereby attracting a highly educated population. However, most of the mpayment users in Mozambique are not university students, and future studies could collect the data in different environments such as markets, companies, and communities, and in a different part of the country or in another African country. This research is related to a single type of technology, m-payment, a comparison with another technology (e.g., m-banking) might reveal other insights and enhance generalization. We used the cultural dimension of uncertainty avoidance, considering that there are other cultural dimensions, it is recommended in future studies to use all the dimensions, with the purpose of bringing more insight. Our models were proposed for and validated in Mozambique, and to generalize the applicability of the study, and to fully represent all potential m-payments users in the African context the samples for future research could be conducted in other African countries. 118 Doctoral Programme in Information Management 119 Doctoral Programme in Information Management Chapter 7 - Bibliography Abbas, H. A., & Hamdy, H. I. (2015). Determinants of continuance intention factor in Kuwait communication market: Case study of Zain-Kuwait. Computers in Human Behavior, 49, 648–657. https://doi.org/10.1016/j.chb.2015.03.035 Afshan, S., & Sharif, A. (2016). Acceptance of mobile banking framework in Pakistan. Telematics and Informatics, 33(2), 370–387. https://doi.org/10.1016/j.tele.2015.09.005 Ahmad, W., & Sun, J. (2018). Antecedents of SMMA continuance intention in two culturally diverse countries: An empirical examination. Journal of Global Information Technology Management, 21(1), 45–68. https://doi.org/10.1080/1097198X.2018.1423840 Aiken, L. S., West, S. G., & Reno, R. R. (1991). Multiple regression: Testing and interpreting interactions. Sage. Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In Action Control (pp. 11–39). https://doi.org/10.1007/978-3-642-69746-3_2 Al-Debei, M. M., Al-Lozi, E., & Papazafeiropoulou, A. (2013). Why people keep coming back to Facebook: Explaining and predicting continuance participation from an extended theory of planned behaviour perspective. Decision Support Systems, 55(1), 43–54. https://doi.org/10.1016/j.dss.2012.12.032 Albashrawi, M., & Motiwalla, L. (2019). Privacy and personalization in continued usage intention of mobile banking: An integrative perspective. Information Systems Frontiers, 21(5), 1031–1043. https://doi.org/10.1007/s10796-017-9814-7 Albuquerque, J. P. de, Diniz, E. H., & Cernev, A. K. (2016). Mobile payments: a scoping study of the literature and issues for future research. Information Development, 32(3), 527–553. https://doi.org/10.1177/0266666914557338 Alraimi, K. M., Zo, H., & Ciganek, A. P. (2015). Understanding the MOOCs continuance: The role of openness and reputation. Computers and Education, 80, 28–38. https://doi.org/10.1016/j.compedu.2014.08.006 Amoroso, D., & Lim, R. (2017). The mediating effects of habit on continuance intention. International Journal of Information Management, 37(6), 693–702. https://doi.org/10.1016/j.ijinfomgt.2017.05.003 Asamoah, D., Takieddine, S., & Amedofu, M. (2020). Examining the effect of mobile money transfer (MMT) capabilities on business growth and development impact. Information Technology for Development, 26(1), 146–161. https://doi.org/10.1080/02681102.2019.1599798 Baabdullah, A. M., Abdallah, A., Rana, N. P., & Kizgin, H. (2019). Consumer use of mobile banking (M-Banking) in Saudi Arabia: Towards an integrated model. International Journal of Information Management, 44, 38–52. https://doi.org/10.1016/j.ijinfomgt.2018.09.002 120 Doctoral Programme in Information Management Bankole, F. O., & Bankole, O. O. (2017). The effects of cultural dimension on ICT innovation: Empirical analysis of mobile phone services. Telematics and Informatics, 34(2), 490–505. https://doi.org/10.1016/j.tele.2016.08.004 Baptista, G., & Oliveira, T. (2015). Understanding mobile banking: The unified theory of acceptance and use of technology combined with cultural moderators. Computers in Human Behavior, 50, 418–430. https://doi.org/10.1016/j.chb.2015.04.024 Baptista, G., & Oliveira, T. (2016). A weight and a meta-analysis on mobile banking acceptance research. Computers in Human Behavior, 63, 480–489. https://doi.org/10.1016/j.chb.2016.05.074 Barnes, S., & Böhringer, M. (2011). Modeling use continuance behaviour in microblogging services: The case of twitter. Journal of Computer Information Systems, 51(4), 1–10. https://doi.org/10.1080/08874417.2011.11645496 Barnes, S. J. (2011). Understanding use continuance in virtual worlds: Empirical test of a research model. Information and Management, 48(8), 313–319. https://doi.org/10.1016/j.im.2011.08.004 Batista, C., & Vicente, P. C. (2014). Introducing mobile money in rural Mozambique: Initial evidence from a field experiment. Nova Africa, 1301. https://doi.org/10.2139/ssrn.2384561 Batista, C., & Vicente, P. C. (2018). Improving access to savings through mobile money: Experimental evidence from smallholder farmers in Mozambique. Nova Africa, 1705. Bayramusta, M., & Nasir, V. A. (2016). A fad or future of IT?: A comprehensive literature review on the cloud computing research. International Journal of Information Management, 36(4), 635–644. https://doi.org/10.1016/j.ijinfomgt.2016.04.006 Belanche, D., Casaló, L. V., Flavián, C., & Schepers, J. (2014). Trust transfer in the continued usage of public e-services. Information and Management, 51(6), 627– 640. https://doi.org/10.1016/j.im.2014.05.016 Bhattacherjee, A., & Barfar, A. (2011). Information Technology Continuance Research: Current State and Future Directions. Asia Pacific Journal of Information Systems, 21(2), 1–18. https://doi.org/10.2307/3250921 Bhattacherjee, A., & Lin, C.-P. (2014). A unified model of IT continuance: Three complementary perspectives and crossover effects. European Journal of Information Systems, 24(4), 1–10. https://doi.org/10.1057/ejis.2013.36 Bhattacherjee, A., & Premkumar, G. (2004). Understanding changes in belief and attitude toward information technology usage: A theoretical model and longitudinal test. MIS Quaterly, 28(2), 229–254. https://doi.org/10.2307/25148634 Bhattacherjee, A. (2001). Understanding information systems continuance: An 121 Doctoral Programme in Information Management expectation confirmation model. MIS Quarterly, 25(3), 351–370. https://doi.org/10.2307/3250921 Bhattacherjee, Anol, Perols, J., & Sanford, C. (2008). Information technology continuance: A theoretic extension and empirical test. Journal of Computer Information Systems, 49(1), 17–26. https://doi.org/10.1080/08874417.2008.11645302 Bøe, T., Gulbrandsen, B., & Sørebø, O. (2015). How to stimulate the continued use of ICT in higher education: Integrating information systems continuance theory and agency theory. Computers in Human Behavior, 50, 375–384. https://doi.org/10.1016/j.chb.2015.03.084 Brislin, R. W. (1970). Back-translation for cross-cultural research. Journal of CrossCultural Psychology, 1(3), 185–216. https://doi.org/10.1177/135910457000100301 Budiardjo, E. K., Pamenan, G., Hidayanto, A. N., Meyliana, & Cofriyanti, E. (2017). The impact of knowledge management system quality on the usage continuity and recommendation intention. Knowledge Management & E-Learning, 9(2), 200– 224. Budner, P., Fischer, M., Rosenkranz, C., Basten, D., & Terlecki, L. (2017). Information system continuance intention in the context of network effects and freemium business models: A replication study of cloud services in Germany. AIS Transactions on Replication Research, 3(December), 1–13. https://doi.org/10.17705/1atrr.00019 Carillo, K., Scornavacca, E., & Za, S. (2015). The role of media dependency in predicting continuance intention to use ubiquitous media systems. Information and Management, 54(3), 317–335. https://doi.org/10.1016/j.im.2016.09.002 Carillo, K., Scornavacca, E., & Za, S. (2017). The role of media dependency in predicting continuance intention to use ubiquitous media systems. Information and Management, 54(3), 317–335. https://doi.org/10.1016/j.im.2016.09.002 Chang, I.-C., Liu, C.-C., & Chen, K. (2014). The effects of hedonic/utilitarian expectations and social influence on continuance intention to play online games. Internet Research, 24(1), 21–45. https://doi.org/10.1108/IntR-02-2012-0025 Chang, Y. P., & Zhu, D. H. (2012). The role of perceived social capital and flow experience in building users’ continuance intention to social networking sites in China. Computers in Human Behavior, 28(3), 995–1001. https://doi.org/10.1016/j.chb.2012.01.001 Chen, C. P., Lai, H. M., & Ho, C. Y. (2015). Why do teachers continue to use teaching blogs? the roles of perceived voluntariness and habit. Computers and Education, 82(1), 236–249. https://doi.org/10.1016/j.compedu.2014.11.017 Chen, I. Y. L. (2007). The factors influencing members’ continuance intentions in professional virtual communities a longitudinal study. Journal of Information 128 Doctoral Programme in Information Management Ifinedo, P. (2017). Examining students’ intention to continue using blogs for learning: Perspectives from technology acceptance, motivational, and social-cognitive frameworks. Computers in Human Behavior, 72, 189–199. https://doi.org/10.1016/j.chb.2016.12.049 Iman, N. (2018). Is mobile payment still relevant in the fintech era? Electronic Commerce Research and Applications, 30(05), 72–82. https://doi.org/10.1016/j.elerap.2018.05.009 INE. (2019). Resultados definitivos do VI recenseamento geral da população e habitação 2017. Instituto Nacional de Estatística (INE). http://www.ine.gov.mz/ Jack, W., & Suri, T. (2011). Mobile money: The economics of M-PESA. Jeyaraj, A., Rottman, J. W., & Lacity, M. C. (2006). A review of the predictors, linkages, and biases in IT innovation adoption research. Journal of Information Technology, 21(1), 1–23. https://doi.org/10.1057/palgrave.jit.2000056 Jin, X. L., Lee, M. K. O., & Cheung, C. M. K. (2010). Predicting continuance in online communities: Model development and empirical test. Behaviour and Information Technology, 29(4), 383–394. https://doi.org/10.1080/01449290903398190 Johns, G. (2006). The essential impact of context on organizational behavior. Academy of Management Journal, 31(2), 386–408. https://doi.org/10.5465/amr.2006.20208687 Johnson, R. E., Rosen, C. C., & Djurdjevic, E. (2011). Assessing the impact of common method variance on higher order multidimensional constructs. Journal of Applied Psychology, 96(4), 744–761. https://doi.org/10.1037/a0021504 Joo, Y. J., Kim, N., & Kim, N. H. (2016). Factors predicting online university students’ use of a mobile learning management system. Educational Technology Research and Development, 64(4), 611–630. https://doi.org/10.1007/s11423-016-9436-7 Joo, Y. J., So, H. J., & Kim, N. H. (2018). Examination of relationships among students’ self-determination, technology acceptance, satisfaction, and continuance intention to use K-MOOCs. Computers and Education, 122(01), 260–272. https://doi.org/10.1016/j.compedu.2018.01.003 Kaewkitipong, L., Chen, C. C., & Ractham, P. (2016). Using social media to enrich information systems field trip experiences: Students’ satisfaction and continuance intentions. Computers in Human Behavior, 63, 256–263. https://doi.org/10.1016/j.chb.2016.05.030 Kang, S. (2014). Factors influencing intention of mobile application use. International Journal of Mobile Communications, 12(4), 360–379. https://doi.org/10.1504/IJMC.2014.063653 Kang, Y. J., & Lee, W. J. (2015). Self-customization of online service environments by users and its effect on their continuance intention. Service Business, 9(2), 321– 342. https://doi.org/10.1007/s11628-014-0229-y 129 Doctoral Programme in Information Management Kari, T., Salo, M., & Frank, L. (2020). Role of situational context in use continuance after critical exergaming incidents. Information Systems Journal, 30(3), 596–633. https://doi.org/10.1111/isj.12273 Karjaluoto, H., Shaikh, A. A., Saarijärvi, H., & Saraniemi, S. (2019). How perceived value drives the use of mobile financial services apps. International Journal of Information Management, 47, 252–261. https://doi.org/10.1016/j.ijinfomgt.2018.08.014 Kaur, P., Dhir, A., Singh, N., Sahu, G., & Almotairi, M. (2020). An innovation resistance theory perspective on mobile payment solutions. Journal of Retailing and Consumer Services, 55(June 2019), 102059. https://doi.org/10.1016/j.jretconser.2020.102059 Khalilzadeh, J., Ozturk, A. B., & Bilgihan, A. (2017). Security-related factors in extended UTAUT model for NFC based mobile payment in the restaurant industry. Computers in Human Behavior, 70, 460–474. https://doi.org/10.1016/j.chb.2017.01.001 Kim, B. (2010). An empirical investigation of mobile data service continuance: Incorporating the theory of planned behavior into the expectation-confirmation model. Expert Systems with Applications, 37(10), 7033–7039. https://doi.org/10.1016/j.eswa.2010.03.015 Kim, B. (2011). Understanding antecedents of continuance intention in socialnetworking services. Cyberpsychology, Behavior and Social Networking, 14(4), 199–205. https://doi.org/10.1089/cyber.2010.0009 Kim, G., Shin, B., & Lee, H. G. (2009). Understanding dynamics between initial trust and usage intentions of mobile banking. Information Systems Journal, 19(3), 283– 311. https://doi.org/10.1111/j.1365-2575.2007.00269.x Kim, Y., & Peterson, R. A. (2017). A Meta-analysis of online trust relationships in Ecommerce. Journal of Interactive Marketing, 38, 44–54. https://doi.org/10.1016/j.intmar.2017.01.001 Koksal, M. H. (2016). The intentions of Lebanese consumers to adopt mobile banking. International Journal of Bank Marketing, 34(3), 327–346. https://doi.org/10.1108/IJBM-03-2015-0025 Koloseni, D., & Mandari, H. (2017). Why mobile money users keep increasing? investigating the continuance usage of mobile money services in Tanzania. 26(2), 117–145. Krasnova, H., Veltri, N. F., Eling, N., & Buxmann, P. (2017). Why men and women continue to use social networking sites: The role of gender differences. Journal of Strategic Information Systems, 26(4), 261–284. https://doi.org/10.1016/j.jsis.2017.01.004 Kroeber, A. L., & Parsons, T. (1958). The concepts of culture and of social system. American Sociological Review, 23(5), 582–583. 130 Doctoral Programme in Information Management Ku, E. C. S., & Chen, C.-D. (2015). Cultivating travellers’ revisit intention to e-tourism service: The moderating effect of website interactivity. Behaviour and Information Technology, 34(5), 465–478. https://doi.org/10.1080/0144929X.2014.978376 Kujala, S., Mugge, R., & Miron-Shatz, T. (2017). The role of expectations in service evaluation: A longitudinal study of a proximity mobile payment service. International Journal of Human Computer Studies, 98, 51–61. https://doi.org/10.1016/j.ijhcs.2016.09.011 Lagna, A., & Ravishankar, M. N. (2021). Making the world a better place with fintech research. Information Systems Journal, 1–42. https://doi.org/10.1111/isj.12333 Lankton, N., & McKnight, H. (2012). Examining two expectation disconfirmation theory models: Assimilation and asymmetry effects. Journal of the Association for Information Systems, 13(2), 88–115. Larsen, T. J., Sørebø, A. M., & Sørebø, Ø. (2009). The role of task-technology fit as users’ motivation to continue information system use. Computers in Human Behavior, 25(3), 778–784. https://doi.org/10.1016/j.chb.2009.02.006 Laukkanen, T. (2007). Internet vs mobile banking: comparing customer value perceptions. Business Process Management Journal, 13(6), 788–797. https://doi.org/10.1108/14637150710834550 Lee, I., Choi, B., Kim, J., & Hong, S.-J. (2007a). Culture-technology fit: Effects of cultural characteristics on the post-adoption beliefs of mobile internet users. International Journal of Electronic Commerce, 11(4), 11–51. https://doi.org/10.2753/JEC1086-4415110401 Lee, I., Choi, B., Kim, J., & Hong, S.-J. (2007b). Culture-Technology Fit: Effects of Cultural Characteristics on the Post-Adoption Beliefs of Mobile Internet Users. International Journal of Electronic Commerce, 11(4), 11–51. https://doi.org/10.2753/JEC1086-4415110401 Lee, J. K., Park, J., Gregor, S., & Yoon, V. (2021). Axiomatic theories and improving the relevance of information systems research. Information Systems Journal, 32(1), 147–171. https://doi.org/10.1287/isre.2020.0958 Lee, M.-C. (2010). Explaining and predicting users’ continuance intention toward elearning: An extension of the expectation-confirmation model. Computers and Education, 54(2), 506–516. https://doi.org/10.1016/j.compedu.2009.09.002 Lee, Y., & Kwon, O. (2011). Intimacy, familiarity and continuance intention: An extended expectation-confirmation model in web-based services. Electronic Commerce Research and Applications, 10(3), 342–357. https://doi.org/10.1016/j.elerap.2010.11.005 Lehto, T., & Oinas-Kukkonen, H. (2015). Explaining and predicting perceived effectiveness and use continuance intention of a behaviour change support system for weight loss. Behaviour & Information Technology, 34(2), 176–189. https://doi.org/10.1080/0144929X.2013.866162 131 Doctoral Programme in Information Management Leidner, D. E., & Kayworth, T. (2006). Review: A review of culture in information systems research: Toward a theory of information technology culture conflict. MIS Quarterly, 30(2), 357–399. Liang, T.-P., Ling, Y.-L., Yeh, Y.-H., & Lin, B. (2013). Contextual factors and continuance intention of mobile services. International Journal of Mobile Communications, 11(4), 313–329. https://doi.org/10.1504/IJMC.2013.055746 Liao, C., Palvia, P., & Lin, H.-N. (2006). The roles of habit and website quality in ecommerce. International Journal of Information Management, 26(6), 469–483. https://doi.org/10.1016/j.ijinfomgt.2006.09.001 Liao, Z., & Shi, X. (2017). Web functionality, web content, information security, and online tourism service continuance. Journal of Retailing and Consumer Services, 39, 258–263. https://doi.org/10.1016/j.jretconser.2017.06.003 Liébana-Cabanillas, F., & Lara-Rubio, J. (2017). Predictive and explanatory modeling regarding adoption of mobile payment systems. Technological Forecasting and Social Change, 120, 32–40. https://doi.org/10.1016/j.techfore.2017.04.002 Liébana-cabanillas, F., Sánchez-Fernández, J., & Muñoz-Leiva, F. (2014). The moderating effect of experience in the adoption of mobile payment tools in virtual social networks: The m-payment acceptance model in virtual social networks (MPAM-VSN). International Journal of Information Management, 34, 151–166. https://doi.org/10.1016/j.ijinfomgt.2013.12.006 Liébana-cabanillas, F., Sánchez-Fernández, J., & Muñoz-Leiva, F. (2018). A global approach to the analysis of user behavior in mobile payment systems in the new electronic environment. Service Business, 12(1), 25–64. https://doi.org/10.1007/s11628-017-0336-7 Liébana-Cabanillas, Francisco, Molinillo, S., & Ruiz-Montañez, M. (2019). To use or not to use, that is the question: Analysis of the determining factors for using NFC mobile payment systems in public transportation. Technological Forecasting and Social Change, 139, 266–276. https://doi.org/10.1016/j.techfore.2018.11.012 Liébana-Cabanillas, Francisco, Singh, N., Kalinic, Z., & Carvajal-Trujillo, E. (2021). Examining the determinants of continuance intention to use and the moderating effect of the gender and age of users of NFC mobile payments: a multi-analytical approach. Information Technology and Management, 22(2), 133–161. https://doi.org/10.1007/s10799-021-00328-6 Limayem, M., & Cheung, C. M. K. (2008). Understanding information systems continuance: The case of Internet-based learning technologies. Information and Management, 45(4), 227–232. https://doi.org/10.1016/j.im.2008.02.005 Limayem, M., & Cheung, C. M. K. (2011). Predicting the continued use of Internetbased learning technologies: the role of habit. Behaviour & Information Technology, 30(1), 91–99. https://doi.org/10.1080/0144929X.2010.490956 Limayem, M., Hirt, S. G., & Cheung, C. M. K. (2007). How habit limits the predictive 132 Doctoral Programme in Information Management power of intention: The case of information systems continuance. MIS Quarterly, 31(4), 705–737. Lin, C.-P., & Bhattacherjee, A. (2010). Extending technology usage models to interactive hedonic technologies: A theoretical model and empirical test. Information Systems Journal, 20(2), 163–181. https://doi.org/10.1111/j.13652575.2007.00265.x Lin, H., Fan, W., & Chau, P. Y. K. (2014). Determinants of users’ continuance of social networking sites: A self-regulation perspective. Information and Management, 51(5), 595–603. https://doi.org/10.1016/j.im.2014.03.010 Lin, K.-M. (2016). Understanding undergraduates’ problems from determinants of Facebook continuance intention. Behaviour & Information Technology, 35(9), 693–705. https://doi.org/10.1080/0144929X.2016.1177114 Lin, K. M. (2011). E-Learning continuance intention: Moderating effects of user elearning experience. Computers and Education, 56(2), 515–526. https://doi.org/10.1016/j.compedu.2010.09.017 Lin, K. M., Chen, N. S., & Fang, K. (2011). Understanding e-learning continuance intention: A negative critical incidents perspective. Behaviour and Information Technology, 30(1), 77–89. https://doi.org/10.1080/01449291003752948 Lin, X., Featherman, M., & Sarker, S. (2017). Understanding factors affecting users’ social networking site continuance: A gender difference perspective. Information and Management, 54(3), 383–395. https://doi.org/10.1016/j.im.2016.09.004 Lindell, M. K., & Whitney, D. J. (2001). Accounting for common method variance in cross-sectional research designs. Journal of Applied Psychology, 86(1), 114–121. https://doi.org/10.1037/0021-9010.86.1.114 Lipsey, M., & Wilson, D. (2001). Practical Meta-analysis. Centers for Teaching and Technology - Book Library, 49. Lowry, P. B., Gaskin, J. E., & Moody, G. D. (2015). Proposing the multimotive information systems continuance model ( MISC ) to better explain end-user system evaluations and continuance intentions. Journal of the Association for Information, 16(7), 515–579. Lu, H.-P., & Lee, M.-R. (2011). Experience differences and continuance intention of blog sharing. Behaviour & Information Technology, 31(11), 1–15. https://doi.org/10.1080/0144929X.2011.611822 Lu, J., Liu, C., & Wei, J. (2017). How important are enjoyment and mobility for mobile applications? Journal of Computer Information Systems, 57(1), 1–12. https://doi.org/10.1080/08874417.2016.1181463 Lu, J., Wei, J., Yu, C., & Liu, C. (2017). How do post-usage factors and espoused cultural values impact mobile payment continuation? Behaviour & Information Technology, 36(2), 140–164. https://doi.org/10.1080/0144929X.2016.1208773 133 Doctoral Programme in Information Management Makina, D. (2017). Introduction to the financial services in Africa special issue. African Journal of Economic and Management Studies, 8(1), 2–7. https://doi.org/10.1108/AJEMS-03-2017-149 Mohammadyari, S., & Singh, H. (2015). Understanding the effect of e-learning on individual performance: The role of digital literacy. Computers & Education, 82, 11–25. https://doi.org/https://doi.org/10.1016/j.compedu.2014.10.025 Morosan, C., & DeFranco, A. (2016). It’s about time: Revisiting UTAUT2 to examine consumers’ intentions to use NFC mobile payments in hotels. International Journal of Hospitality Management, 53, 17–29. https://doi.org/10.1016/j.ijhm.2015.11.003 Mouakket, S. (2015). Factors influencing continuance intention to use social network sites: The Facebook case. Computers in Human Behavior, 53, 102–110. https://doi.org/10.1016/j.chb.2015.06.045 Mouakket, S. (2018). The role of personality traits in motivating users’ continuance intention towards Facebook: Gender differences. Journal of High Technology Management Research, 29(1), 124–140. https://doi.org/10.1016/j.hitech.2016.10.003 Muñoz-Leiva, F., Mayo-Muñoz, X., & De la Hoz-Correa, A. (2018). Adoption of homesharing platforms: a cross-cultural study. Journal of Hospitality and Tourism Insights, 1(3), 220–239. https://doi.org/10.1108/jhti-01-2018-0007 Nabavi, A., Taghavi-Fard, M. T., Hanafizadeh, P., & Taghva, M. R. (2016). Information Technology Continuance Intention: A Systematic Literature Review. International Journal of E-Business Research, 12(1), 58–95. https://doi.org/10.4018/IJEBR.2016010104 Nascimento, B., Oliveira, T., & Tam, C. (2018). Wearable technology: What explains continuance intention in smartwatches? Journal of Retailing and Consumer Services, 43, 157–169. https://doi.org/https://doi.org/10.1016/j.jretconser.2018.03.017 Odoom, R., & Kosiba, J. P. (2020). Mobile money usage and continuance intention among micro enterprises in an emerging market – the mediating role of agent credibility. Journal of Systems and Information Technology, 22(4), 97–117. https://doi.org/10.1108/JSIT-03-2019-0062 Oghuma, A. P., Libaque-Saenz, C. F., Wong, S. F., & Chang, Y. (2016). An expectation-confirmation model of continuance intention to use mobile instant messaging. Telematics and Informatics, 33(1), 34–47. https://doi.org/10.1016/j.tele.2015.05.006 Oliveira, T., Alhinho, M., Rita, P., & Dhillon, G. (2017). Modelling and testing consumer trust dimensions in e-commerce. Computers in Human Behavior, 71, 153–164. https://doi.org/10.1016/j.chb.2017.01.050 Oliveira, T., Faria, M., Thomas, M. A., & Popovič, A. (2014). Extending the 134 Doctoral Programme in Information Management 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 Oliveira, T., Thomas, M., Baptista, G., & Campos, F. (2016). Mobile payment: Understanding the determinants of customer adoption and intention to recommend the technology. Computers in Human Behavior, 61, 404–414. https://doi.org/10.1016/j.chb.2016.03.030 Oliveira, T., Thomas, M., & Espadanal, M. (2014). Assessing the determinants of cloud computing adoption: An analysis of the manufacturing and services sectors. Information and Management, 51(5), 497–510. https://doi.org/10.1016/j.im.2014.03.006 Oliver, R. L. (1986). A cognitive model of the antecedents and consequences of satisfaction decisions. Journal of Marketing Research, 17(4), 460–469. https://doi.org/10.2307/3150499 Omigie, N. O., Zo, H., Rho, J. J., & Ciganek, A. P. (2017). Customer pre-adoption choice behavior for M-PESA mobile financial services: Extending the theory of consumption values. Industrial Management and Data Systems, 117(5), 910–926. https://doi.org/10.1108/IMDS-06-2016-0228 Ortigão, M., Macome, E., & Vicente, P. (2015). Electronic payment in Mozambique: A baseline on their adoption in Maputo and Matola. Nova Africa, 1503. Ouyang, Y., Tang, C., Rong, W., Zhang, L., Yin, C., & Xiong, Z. (2017). Tasktechnology fit aware expectation-confirmation model towards understanding of MOOCs continued usage. 50th Hawaii International Conference on System Sciences, 174–183. Pal, A., Herath, T., De’, R., & Rao, H. R. (2020). Contextual facilitators and barriers influencing the continued use of mobile payment services in a developing country: insights from adopters in India. Information Technology for Development, 26(2), 394–420. https://doi.org/10.1080/02681102.2019.1701969 Palvia, P. (2009). The role of trust in e-commerce relational exchange: A unified model. Information and Management, 46(4), 213–220. https://doi.org/10.1016/j.im.2009.02.003 Park, J. K., Ahn, J., Thavisay, T., & Ren, T. (2019). Examining the role of anxiety and social influence in multi-benefits of mobile payment service. Journal of Retailing and Consumer Services, 47, 140–149. https://doi.org/10.1016/j.jretconser.2018.11.015 Park, M., Jun, J., & Park, H. (2017). Understanding mobile payment service continuous use intention: An expectation - Confirmation model and inertia. Quality Innovation Prosperity, 21(3), 78–94. https://doi.org/10.12776/QIP.V21I3.983 Pereira, F. A. de M., Ramos, A. S. M., Andrade, A. P. V. de, & Oliveira, B. M. K. de. (2015). Continued usage of e-learning: Expectations and performance. In Journal of Information Systems and Technology Management (Vol. 12, Issue 2, pp. 333– 135 Doctoral Programme in Information Management 350). scielo. https://doi.org/10.4301/S1807-17752015000200008 Persaud, A., & Azhar, I. (2012). Innovative mobile marketing via smartphones: Are consumers ready? Marketing Intelligence & Planning, 30(4), 418–443. https://doi.org/10.1108/02634501211231883 Pinsonneault, A., & Kraemer, K. L. (1993). Survey research methodology in management information systems: An assessment. Journal of Management Information Systems, 10(2), 75–105. https://doi.org/10.1080/07421222.1993.11518001 Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879–903. https://doi.org/10.1037/0021-9010.88.5.879 Premkumar, G., & Bhattacherjee, A. (2008). Explaining information technology usage: A test of competing models. Omega, 36(1), 64–75. https://doi.org/10.1016/j.omega.2005.12.002 Rahi, S., Khan, M. M., & Alghizzawi, M. (2020). Extension of technology continuance theory (TCT) with task technology fit (TTF) in the context of Internet banking user continuance intention. International Journal of Quality and Reliability Management, ahead-of-p(ahead-of-print). https://doi.org/10.1108/IJQRM-032020-0074 Rahman, S. A., Alam, M. M. D., & Taghizadeh, S. K. (2020). Do mobile financial services ensure the subjective well-being of micro-entrepreneurs? An investigation applying UTAUT2 model. Information Technology for Development, 26(2), 421–444. https://doi.org/10.1080/02681102.2019.1643278 Raman, P., & Aashish, K. (2021). To continue or not to continue: a structural analysis of antecedents of mobile payment systems in India. International Journal of Bank Marketing, 39(2), 242–271. https://doi.org/10.1108/IJBM-04-2020-0167 Rana, N. P., Dwivedi, Y. K., & Williams, M. D. (2015). A meta-analysis of existing research on citizen adoption of e-government. Information Systems Frontiers, 17(3), 547–563. https://doi.org/10.1007/s10796-013-9431-z Rezvani, A., Khosravi, P., & Dong, L. (2017). Motivating users toward continued usage of information systems: Self-determination theory perspective. Computers in Human Behavior, 76, 263–275. https://doi.org/10.1016/j.chb.2017.07.032 Rhee, H. van, Suurmond, R., & Hak, T. (2015). User manual for meta-essentials: Workbooks for meta-analyses. The Netherlands: Erasmus Research Institute of Management, 1–49. Ringle, C. M., Wende, S., & Becker, J.-M. (2015). SmartPLS 3. Bönningstedt: SmartPLS. http://www.smartpls.com Ringle, C. M., Wende, S., & Will, A. (2005). SmartPLS 2.0. M3. Hamburg: SmartPLS. 136 Doctoral Programme in Information Management Roca, J. C., Chiu, C. M., & Martínez, F. J. (2006). Understanding e-learning continuance intention: An extension of the Technology Acceptance Model. International Journal of Human Computer Studies, 64(8), 683–696. https://doi.org/10.1016/j.ijhcs.2006.01.003 Rosenthal, R., & Rubin, D. B. (1982). A simple, general purpose display of magnitude of experimental effect. Journal of Educational Psychology, 74(2), 166–169. https://doi.org/10.1037/0022-0663.74.2.166 Ryans, A. B. (1974). Estimating consumer preferences for a new durable brand in an established product class. Journal of Marketing Research, 11(4), 434–443. https://doi.org/10.2307/3151290 Ryu, H. S. (2018). What makes users willing or hesitant to use Fintech?: the moderating effect of user type. Industrial Management and Data Systems, 118(3), 541–569. https://doi.org/10.1108/IMDS-07-2017-0325 Salehan, M., Kim, D. J., & Lee, J. N. (2018). Are there any relationships between technology and cultural values? A country-level trend study of the association between information communication technology and cultural values. Information and Management, 55(6), 725–745. https://doi.org/10.1016/j.im.2018.03.003 Sällberg, H., & Bengtsson, L. (2016). Computer and smartphone continuance intention: A motivational model. Journal of Computer Information Systems, 56(4), 321–330. https://doi.org/10.1080/08874417.2016.1164007 Santini, F. D. O., Ladeira, W. J., Sampaio, C. H., Perin, M. G., & Dolci, P. C. (2019). A meta-analytical study of technological acceptance in banking contexts. International Journal of Bank Marketing, 37(3), 755–774. https://doi.org/10.1108/IJBM-04-2018-0110 Santini, F. de O., Ladeira, W. J., Sampaio, C. H., Perin, M. G., & Dolci, P. C. (2019). Propensity for technological adoption: an analysis of effects size in the banking sector. Behaviour and Information Technology, 1–15. Schmidt., F. L., & Hunter, J. E. (2004). Methods of meta-analysis: Correcting error and bias in research findings. In Sage. Schmidt, F. L., & Hunter, J. E. (2016). Methods of Meta-Analysis: Correcting Error and Bias in Research Findings. In Sage. Seol, S., Lee, H., Yu, J., & Zo, H. (2016). Continuance usage of corporate SNS pages: A communicative ecology perspective. Information & Management, 53(6), 740– 751. https://doi.org/https://doi.org/10.1016/j.im.2016.02.010 Shaikh, A. A., & Karjaluoto, H. (2015). Making the most of information technology & systems usage: A literature review, framework and future research agenda. Computers in Human Behavior, 49, 541–566. https://doi.org/10.1016/j.chb.2015.03.059 Shao, Z. (2018). Examining the impact mechanism of social psychological motivations on individuals’ continuance intention of MOOCs-The moderating effect of 137 Doctoral Programme in Information Management gender. Internet Research, 28(1), 232–250. https://doi.org/10.1108/IntR-11-20160335 Shao, Z., Zhang, L., Li, X., & Guo, Y. (2019). Antecedents of trust and continuance intention in mobile payment platforms: The moderating effect of gender. Electronic Commerce Research and Applications, 33(11), 100823. https://doi.org/10.1016/j.elerap.2018.100823 Sharma, K. S., & Sharma, M. (2019). Examining the role of trust and quality dimensions in the actual usage of mobile banking services: An empirical investigation. International Journal of Information Management, 44, 65–75. https://doi.org/10.1016/j.ijinfomgt.2018.09.013 Shiau, W. L., Yuan, Y., Pu, X., Ray, S., & Chen, C. C. (2020). Understanding fintech continuance: perspectives from self-efficacy and ECT-IS theories. Industrial Management and Data Systems, 120(9), 1659–1689. https://doi.org/10.1108/IMDS-02-2020-0069 Shin, D. H., & Biocca, F. (2017). Health experience model of personal informatics: The case of a quantified self. Computers in Human Behavior, 69, 62–74. https://doi.org/10.1016/j.chb.2016.12.019 Shin, D. H., Lee, S., & Hwang, Y. (2017). How do credibility and utility play in the user experience of health informatics services? Computers in Human Behavior, 67, 292–302. https://doi.org/10.1016/j.chb.2016.11.007 Singh, N., Sinha, N., & Liébana-cabanillas, F. J. (2020). Determining factors in the adoption and recommendation of mobile wallet services in India : Analysis of the effect of innovativeness, stress to use and social influence. International Journal of Information Management, 50, 191–205. https://doi.org/10.1016/j.ijinfomgt.2019.05.022 Sinha, M., Majra, H., Hutchins, J., & Saxena, R. (2019). Mobile payments in India: The privacy factor. International Journal of Bank Marketing, 37(1), 192–209. https://doi.org/10.1108/IJBM-05-2017-0099 Smith, M. L. (2006). Overcoming theory-practice inconsistencies: Critical realism and information systems research. Information and Organization, 16(3), 191–211. https://doi.org/10.1016/j.infoandorg.2005.10.003 Soffer, P., & Hadar, I. (2007). Applying ontology-based rules to conceptual modeling: A reflection on modeling decision making. European Journal of Information Systems, 16(5), 599–611. https://doi.org/10.1057/palgrave.ejis.3000683 Song, J., Kim, J., & Cho, K. (2017). Understanding users’ continuance intentions to use smart-connected sports products. Sport Management Review, 21(5), 477–490. http://dx.doi.org/10.1016/j.smr.2017.10.004 Sørebø, Ø., Halvari, H., Gulli, V. F., & Kristiansen, R. (2009). The role of selfdetermination theory in explaining teachers’ motivation to continue to use elearning technology. Computers and Education, 53(4), 1177–1187.