An integrated model of mobile internet services usage and

Int. J. Mobile Communications, Vol. 8, No. 4, 2010 411

Copyright © 2010 Inderscience Enterprises Ltd.

An integrated model of mobile internet services usage and continuance

Hee-Woong Kim Graduate School of Information, Yonsei University, Shinchon 134, Seoul 120-749, Korea Fax: +82-2-2123-8654 E-mail: [email protected]

Kee-Young Kwahk* College of Business Administration, Kookmin University, 861-1, Jeongneung-dong, Seongbuk-gu, Seoul 136-702, Korea Fax: +82-2-910-4519 E-mail: [email protected] *Corresponding author

Hyoung-Yong Lee School of Business Administration, Hansung University, 389 Samseon-dong 3-ga, Seongbuk-gu, Seoul 136-792, Korea Fax: +82-2-760-4482 E-mail: [email protected]

Abstract: This study examines the usage and continuance of mobile internet services by developing a theoretical framework incorporating the cognitive, affective and judgemental components, as well as automatic factors, such as habit. Based on this framework, we examine and compare the usage and continuance of mobile internet services by collecting data from mobile service users. By identifying the different influences of the determinants that drive usage behaviour and continuance intention, the study results have helped advance information systems’ theory and offered insights into mobile internet service providers.

Keywords: mobile internet services; mobile communications; structural equation model; usage and continuance.

Reference to this paper should be made as follows: Kim, H.W., Kwahk, K.Y. and Lee, H.Y. (2010) ‘An integrated model of mobile internet services usage and continuance’, Int. J. Mobile Communications, Vol. 8, No. 4, pp.411–429.

412 H.W. Kim et al.

Biographical notes: Hee-Woong Kim is an Assistant Professor in the Graduate School of Information at Yonsei University. Before joining Yonsei University, he worked at the National University of Singapore. His research interests include IT-induced organisational change and the application of IS for the success of online business. His research work has been published in Communications of the ACM, DSS, European Journal of Operational Research, I&M, IJEC, International Journal of Human Computer Studies, Journal of American Society for Information Systems and Technology, Journal of the Association for Information Systems, Journal of Retailing, Psychology and Marketing, and MIS Quarterly.

Kee-Young Kwahk is an Associate Professor at the College of Business Administration of Kookmin University in Seoul, Korea. His research interests include IT-enabled organisational agility, knowledge management, e-business, and social network analysis and its application to IS domain. His research papers have been published and presented in several refereed journals and conferences.

Hyoung-Yong Lee is an Assistant Professor at the Business School at Hansung University, Seoul, Korea. He received his PhD Degree from Korea Advanced Institute of Science and Technology. His current research focuses on customer behaviour on internet, virtual community, and e-commerce.

1 Introduction

According to IDC (www.idc.com), about 1.4 billion people were internet users in 2008. This number is expected to surpass 1.9 billion in 2012. Roughly, 40% of all internet users currently have mobile internet access. The number of mobile internet service users reached 56 million in 2008 and is forecasted to surpass 1.5 billion in 2012. The most popular online activities among mobile internet users include searching the web, accessing news and sports information, using instant messaging, using internet e-mails and downloading music, videos and ringtones (Safko and Brake, 2009). Mobile communication and services have been widely adopted and used worldwide for different purposes, e.g., student usage in Greek (Economides and Grousopoulou, 2008), Wi-Fi services in German public hot spots (Husig and Hipp, 2009a, 2009b), 3G mobile phones in China (Liu et al., 2008), location-based services in Singapore (Xu et al., 2009) and wireless internet services in Malaysia (Parveen et al., 2009). Mobile service managers recognise that achieving strong and sustained customers is crucial (Lin and Shih, 2008). Therefore, research into mobile service usage has recently emerged as a practically important issue in IS literature (Kivi, 2009; Liao et al., 2007; Limayem et al., 2007; Nickerson et al., 2008).

A great number of previous studies (Lee and Jun, 2007; Li and McQueen, 2008; Lin and Liu, 2009; Liu et al., 2008; Xu and Yuan, 2009) were devoted to the topic of the users’ adoption of mobile services. The initial usage is an important measure of mobile services success, but it does not necessarily result in the desired managerial performance unless the usage is continuous (Bhattacherjee, 2001b). In the context of mobile communications, many studies have focused on the adoption and use of mobile internet services because the services and the relevant technologies are relatively new. A few studies (Hung et al., 2007; Lin and Shih, 2008) have begun to examine the continuance of

An integrated model of mobile internet services usage and continuance 413

mobile internet services. The change behind this shift in research trends from the initial adoption to the continuance usage stems from the realisation that mere acceptance will not attract future IT industry investment.

Further, usage experiences of mobile internet services incorporate diverse aspects such as cognitive, affective and judgemental because most users use mobile internet services for personal purposes, and they have to pay usage fees. In contrast, company employees use enterprise Information Systems (IS) for task-related purposes, and they do not have to pay for the usage. Mobile service users would be, therefore, salient to affective aspect (e.g., emotion or playfulness) and judgemental ones (e.g., the money value of the mobile services) as well as cognitive ones (e.g., usefulness). Those diverse aspects would affect the way the users make decisions about the usage and continuance intention of mobile services. However, few studies have examined the diverse aspects using the same research model. The Theory of Planned Behaviour (TPB) (Ajzen, 1991) and the Technology Acceptance Model (TAM) (Davis, 1989) have only considered the cognitive aspects in examining human behaviour, especially in IS usage.

Given this motivation, the primary objective of this study is to examine and compare mobile internet services usage and continuance by developing an integrated model. This study develops an integrated theoretical model based on cognitive, affective and judgemental components, as well as automatic components such as habit and it examines their effects on the usage and continuance intention by collecting data from mobile internet services users. We expect this study to make important theoretical and practical contributions to the studies of mobile services usage and the continuance intention. On the theoretical side, we propose an integrated model explaining both usage behaviour and continuance intention. This model highlights the different decision-making processes when users form usage intentions and how they actually use mobile internet services. This paper also shows the discrepancy between behaviour and intention in the context of mobile internet services. In practical terms, this study identifies different key drivers of mobile internet services usage and the continuance intention; thereby, providing a set of tangible guidelines that help the mobile internet service providers motivates usage.

2 Theoretical framework

In the past two decades of IS usage research, there has been much focus on cognitive behavioural models, including TPB and TAM. However, the cognitive behavioural models have been criticised for being incomplete due to their ignorance of other factors affecting human behaviour. An emerging approach to consumer behaviour research has pursued a more reactive view of consumption that emphasises the role of emotional responses in which both conscious and unconscious thoughts and feelings surround and sustain the consumption context. Moreover, some studies showed that frequently performed behaviours tend to become habitual, and thus automatic over time, and this adds to the explanatory power of IS usage behaviour (Triandis, 1980; Limayem et al., 2003). Based on an in-depth literature review and the characteristics of mobile services usage, we conclude that our conceptual framework consists of cognition (Lee and Ahn, 2009), affect (Belk, 1985; Holbrook et al., 1990; Kang et al., 2009), judgement (Ajzen, 1991) and habit (Ouellette and Wood, 1998), as is shown in Figure 1.

414 H.W. Kim et al.

Figure 1 Theoretical framework

2.1 Judgement

Judgement is the conscious evaluation of the potential consequences of behaviour based on both cognition and affect (Bagozzi et al., 1999). Overall judgment is the combination of both cognitive judgment and affective judgement. The cognitive element consists of the subjects’ judgments of some beliefs, whereas the affective element is based on the feelings that the subjects experienced. Holbrook et al. (1990) conceptualised affective judgement as the appreciation that represents an affective evaluation and cognitive judgement as the reasons that represent a cognitive evaluation. In the context of mobile internet services, users may judge the value of the mobile services. Especially, users would judge whether or not the mobile internet services are worth the usage fee. Judgement then plays a key role in human decision making such as deciding mobile internet services usage and continuance (Ajzen, 1991; Kim et al., 2006, 2007).

2.2 Feeling

The affective experience is an important dimension of the whole usage experience in the context of mobile internet services since many people use mobile internet services for personal purposes (e.g., downloading music, videos and ringtones and using instant messaging). The affect towards a behaviour reflects the direct emotional response to the thought of the behaviour. Consumer research (Derbaix and Pham, 1991) and social psychology (Zajonc, 1980) have proposed that affect plays a central role in the decision-making process. The feeling-as-information model further reinforces the relationship between feeling and judgement (Schwarz, 2000). This model holds that judgements are based on perceptible feelings. The feeling-as-information model also argues that feeling helps to form overall judgment through a controlled inferential process (Schwarz, 2000). Therefore, feeling is considered as another antecedent of judgement.

Regarding the feeling-behaviour (intention) link, consumer behaviour (Derbaix and Pham, 1991) and social psychology (Zajonc, 1980) research proposed that most behavioural sequences generally contain both affective and cognitive components. Romer (2000) sets forth the idea of the hybrid mechanism as one of the human behaviour feeling systems that uses both belief-based and feeling-based mechanisms. When people employ a belief-based mechanism, they compute the outcome function of each action using its realisation probability before making a decision. In contrast, when employing a feeling-based mechanism, they will be conscious of the hedonic state generated, and they choose an action that offers a higher value at the given hedonic state. These imply feelings are natural antecedents of behaviour and intention.


2.3 Belief

In the context of mobile internet services, people perceive several characteristics of focal services such as usefulness. Especially, mobile internet services are characterised by their ubiquitousness. That is, people can access the internet and get information on the road anytime and anywhere. Belief refers to the perception of the cognitive characteristics of mobile services. Thus, people would judge mobile internet services based on perception of their characteristics, and belief in their rational decision-making (Holbrook et al., 1990).

Many previous studies also viewed belief as having a direct influence on behaviour, implying a direct belief-behaviour (intention) linkage. The cognitive factors of behavioural intention were initially raised by Triandis (1980), who posited cognition as having a direct effect on behavioural intention. Later, Davis (1989) incorporated the straight effect of beliefs on intention in TAM.

2.4 Habit

Habits are commonly understood as learned sequences of acts that become automatic responses to specific situations that may be functional in achieving certain goals or reaching end states. When a behaviour is repeatedly executed (in stable contexts), performing this activity is no longer a rational decision-making process (Verplanken et al., 1998), and the behaviour becomes a habitual one. In the context of mobile internet services, many users repeatedly send instant messages, search the web and access the news on the road and during transportation. It has been shown that measuring past behaviour enables the prediction of future behaviour, over and beyond measures of judgement, via a non-conscious process that is usually termed habit (Sheeran and Orbell, 1999). According to this view, behaviours that have been performed many times in the past are encoded in the memory such that environmental cues serve to automatically elicit the behaviour. Thus, behaviour may be guided more by the automaticity of the stimulus-response relation, and less by judgement (Verplanken et al., 1998).

3 Research model and hypotheses

We developed a research model based on theoretical framework. We identified perceived value as a construct representing judgement, usefulness as a construct representing belief and emotion as a construct representing feeling, as is shown in Figure 2.

Figure 2 Research model

416 H.W. Kim et al.

3.1 Perceived value

According to the economic theory of utility, customers try to achieve maximum utility or satisfaction, given their resource limitations. The definition of perceived value reflects this by comparing benefits with sacrifices. Zeithaml (1988) explains perceived value as the consumers’ perceptions of what is received and what is given, which determine the consumer’s overall assessment of the utility of a product. This study refers to the perceived value of mobile internet services as a consumer’s overall judgement of mobile internet services based on the considerations of its benefits and sacrifices needed to acquire and/or use them following previous research (Kim et al., 2006). Perceived value, in certain aspects, is similar to attitude in TPB (Ajzen, 1991) that explains attitude as a judgement that determines behaviour and intention. The majority of studies on IT adoption have focused on the relationship between attitude (judgement) and usage intention. Karahanna et al. (1999) attempted to test employees’ attitudes of their behavioural intention to continue to use Windows OS software and found a significant relationship between the two constructs. Hence, we propose the following hypotheses:

H1a: The perceived value of mobile internet services is positively related to the usage.

H1b: The perceived value of mobile internet services is positively related to the continuance intention.

3.2 Usefulness

Usefulness means the degree to which a person believes that using a particular system will be advantageous in performing his or her tasks (Davis, 1989). For this study, we define a task as an activity or function to be performed using the mobile internet, such as reading news, checking e-mails, reserving movie tickets and getting the latest lottery results. We can consider usefulness and perceived value as measures of expectancy and judgement, respectively. Thus, the relationship between usefulness and perceived value is supported by the expectancy value theory, which suggests that people guide themselves according to their expectations and evaluations (Palmgreen, 1984). Therefore, if a technology performs up to expectation, provides gains over alternative services and helps users in difficult situations (i.e., it is useful), then users are likely to favourably evaluate the use of the technology. Mobile internet is more appealing to customers as it continues to be improved to perform tasks conveniently, effectively and efficiently; particularly, when the customer is travelling. Hence, we propose the following hypothesis:

H2: The usefulness of mobile internet services is positively related to their perceived values.

Individuals adopt technology because they derive some benefits from its use and they consequently feel it to be useful. Davis (1989) has shown that perceived usefulness affects usage behaviour of IT, possibly because people will use a specific IT only if they perceive that such usage will help them achieve the desired task performance. Hence, in the context of the mobile internet, we can postulate that perceived usefulness is positively related to mobile internet usage. Besides, usefulness also affects continuance intention of mobile internet usage at the post-adoption stage. Continuance intention of


mobile internet usage is defined as the actual intention to continue using mobile internet services at the post-adoption stage (Bhattacherjee, 2001a). If an individual believes that continuing to use mobile internet services will help in attaining a certain goal (i.e., being useful), then he or she will want to continue to use them. Bhattacherjee (2001a) found that perceived usefulness has a strong and positive impact on continuance intention. Therefore, customers may continue to use mobile internet services if they believe that using it will help to gain some desired benefits. Hence, we propose the following hypotheses:

H3a: The usefulness of mobile internet services is positively related to the usage.

H3b: The usefulness of mobile internet services is positively related to the continuance intention.

3.3 Emotion

As one category of affect, Bagozzi et al. (1999) defined emotion as a mental state of readiness that arises from the cognitive appraisals of events or thoughts and it is often physically expressed (e.g., facial features). Emotions play an important role in people’s evaluation processes. A competing explanation based on the feeling-as-information model (Schwarz, 2000) suggests that individuals may assume that their emotions are affective reactions to the object being evaluated, and thus base their evaluations on their affective states. Therefore, people may use their emotions as the basis of evaluating objects. Gardner (1985) has shown that affective responses are able to influence cognitive processes such as evaluation, recall and judgement. Furthermore, the emotion-perceived value relationship is also supported by the mental accounting theory (Thaler, 1985), which states that the acquisition utility consists of both hedonic and utilitarian components in which emotion belongs to the hedonic component. Hence, we propose the following hypothesis:

H4: The emotion generated from the usage of mobile internet services is positively related to the perceived value.

Emotion can have a direct influence on behaviour. Lazarus (1991) explains that emotion can indeed be represented as a unique and unmediated antecedent of behaviour or behavioural intention. Consistent with this, Allen et al. (1992) found that emotion can have a direct influence on behaviour. In the context of mobile internet services, individuals may be willing to use mobile internet if they experience immediate pleasure or joy. Likewise, emotion can also influence the continuance intention of mobile internet usage at the post-adoption stage. Therefore, we propose the following hypotheses:

H5a: The emotion generated from the usage of mobile internet services is positively related to the usage.

H5b: The emotion generated from the usage of mobile internet services is positively related to the continuance intention.

3.4 Habit

Habits reflect automatic behaviour tendencies developed during the individual’s history, such that particular stimuli elicit the behaviour even when the individual does not instruct

418 H.W. Kim et al.

himself to perform it (Limayem et al., 2003). Aarts et al. (1998) proposed that behaviour is guided by automated cognitive processes when repeatedly and habitually performed, rather than being preceded by elaborated decision processes (i.e., a decision based on judgement). The habit-behaviour relationship is also supported by Triandis’ model (Triandis, 1980) in which habit directly influences behaviour. As individuals get used to using mobile internet services, they probably continue to habitually use it. Therefore, we propose the following hypotheses:

H6a: The habit of using mobile internet services is positively related to the usage.

H6b: The habit of using mobile internet services is positively related to continuance intention.

4 Research methodology

4.1 Data collection procedure and sample characteristics

Empirical data for this study were collected via an internet survey after which further data were collected by quota sampling that only included respondents above 25 years old, through face-to-face and mail questionnaire surveys. E-mails were sent out via the university’s e-mailing list to the undergraduates and graduates of a major university in Singapore inviting them to take the survey based on whether or not they have previously used mobile internet services. To improve the response rate, S$50 was offered to 20 respondents who were randomly selected for the prize. The returned questionnaires were screened for completeness and reliability, and 290 responses were found to be complete and usable.

Among the detailed descriptive statistics of the respondents’ characteristics shown in Table 1, 67.4% of the respondents were male and 77.9% were 20–29 years old, which is a similar population distribution to that used in previous studies conducted in Singapore (Kim et al., 2006). Because a university mailing list was used, 63.4% were university student respondents. The high penetration rate of mobile phones in Singapore was reflected in the percentage of users, 89.3%, using mobile phones to access mobile internet services.

Table 1 Demographic data of the respondents

Respondent profiles Frequency Percentage (%)

Male 196 67.6 Gender Female 94 32.4 <20 10 3.4 20–24 164 56.6 25–29 52 17.9 30–34 29 10.0 35–39 11 3.8

Age

40> 24 8.3


Table 1 Demographic data of the respondents (continued)

Respondent profiles Frequency Percentage (%)

Graduate 30 10.3 Undergraduate 154 53.1 Junior College/Polytechnic student 3 1.0 Self-employed 10 3.4 Professional 64 22.1 Unemployed 5 1.7 Others 22 7.6

Profession

No answer 2 0.7 Mobile phone 259 89.3 PDA 16 5.5

Mobile device

PDA phone 15 5.2 Total 290 100.0

4.2 Measurement development

We developed our data collection instrument by adopting existing validated questions wherever possible. We adopted the scales from Davis (1989) to measure usefulness. Emotion items were adopted from Cheung et al. (2000) and from Cohen and Areni (1991). We adopted the scales for the perceived value from Sirdeshmukh et al. (2002). Continuance intention scales were adopted from Bhattacherjee (2001b), while usage scales were adopted from Cheung et al. (2000). The items for habit were adopted from Limayem and Hirt (2003). All question items were measured with a seven-point, Likert-type scale, with anchors ranging from ‘strongly disagree’ to ‘strongly agree’. Two IS researchers and five scholars reviewed the instrument and checked the face validity of the items. As a pre-test, the questionnaires were discussed during interviews conducted with 10 people. We obtained feedback about the instrument content, question ambiguity, scale format, length and the overall format and design of the questionnaire. The final instrument used for data collection is presented in Appendix.

5 Data analysis and results

5.1 Confirmatory Factor Analysis (CFA)

We conducted the Confirmatory Factor Analysis (CFA) by creating a LISREL path diagram. Following the recommended methodological procedures (Gefen et al., 2000), the measurement model in CFA was revised by sequentially dropping items sharing a high degree of residual variance with other items in order to purge items that violate unidimensionality. For both models, we dropped two items: the second (HAB2) item of habit since it shares a high degree of residual variance with EMO3, EMO4, HAB4, HAB5, USE1, USE2 and USE3. After HAB2 was dropped, CFA showed a very good model fit: Goodness-of-Fit Index (GFI) = 0.87, Normed Fit Index (NFI) = 0.96, Comparative Fit Index (CFI) = 0.98, Adjusted Goodness-of-Fit Index (AGFI) = 0.83 and

420 H.W. Kim et al.

Root Mean Square Error of Approximation (RMSEA) = 0.073. We consistently dropped the same item HAB2 for the continuance intention model. After dropping those items, CFA showed an acceptable model fit: GFI = 0.88, NFI = 0.97, CFI = 0.98, AGFI = 0.85 and RMSEA = 0.067. The fourth item of INT (INT4) was also dropped because it had a very low value of standard loading.

We next tested the convergent validity. There are three criteria for assessing the convergent validity. First, standardised path loadings, which indicate the degree of association between the underlying variable factor and each item, should be greater than 0.7 and statistically significant (Gefen et al., 2000). Second, the composite reliabilities, as well as Cronbach’s alphas, should be higher than 0.7. Third, the Average Variance Extracted (AVE) for each factor should exceed 0.5. The figures in Table 2 show that all path loadings were greater than 0.7 and were statistically significant. All the reliability measures were above 0.7, and all the AVE values were above 0.5. Thus, the convergent validity of all the items has been established.

Table 2 Result of convergent validity testing

Items Std loading Std error t-value AVE CR Alpha

USF1 0.84 0.29 17.38 USF2 0.87 0.24 18.48 USF3 0.89 0.21 19.01 USF4 0.89 0.21 19.08

0.762 0.928 0.9279

EMO1 0.85 0.27 17.91 EMO2 0.89 0.21 18.96 EMO3 0.91 0.16 20.10 EMO4 0.85 0.27 17.85 EMO5 0.72 0.48 13.96

0.721 0.928 0.9256

HAB1 0.74 0.45 14.10 HAB3 0.69 0.53 12.76 HAB4 0.82 0.34 16.20 HAB5 0.83 0.31 16.56

0.594 0.853 0.8503

VAL1 0.75 0.43 14.69 VAL2 0.80 0.37 15.95 VAL3 0.88 0.22 18.67 VAL4 0.91 0.18 19.55

0.700 0.903 0.9016

USG1 0.87 0.24 18.50 USG2 0.72 0.49 13.82 USG3 0.87 0.25 18.26 USG4 0.92 0.15 20.18

0.718 0.910 0.9088

INT1 0.95 0.11 21.44 INT2 0.92 0.15 20.60 INT3 0.93 0.14 20.72

0.908 0.974 0.8919


We assessed the discriminant validity by comparing the square root of AVE of each construct with the correlations between that construct and other constructs. As shown in Table 3, the square root of AVE of each construct exceeded the correlations between that construct and other constructs. Hence, the questions used in this study had discriminant validity.

Table 3 The results of discriminant validity testing and correlations

Variables Mean SD USF EMO VAL HAB USG INT

USF 4.54 1.12 0.87 EMO 4.51 1.13 0.47* 0.85 VAL 4.03 1.17 0.50* 0.61* 0.84 HAB 3.46 1.32 0.47* 0.43* 0.49* 0.77 USG 3.22 1.35 0.42* 0.32* 0.42* 0.66* 0.85 INT 4.48 0.90 0.47* 0.30* 0.40* 0.43* 0.46* 0.95

The diagonal values are the square roots of the Average Variance Extracted for each construct (AVE). *Correlation is significant at the 0.01 level (2-tailed).

5.2 Hypothesis testing

After establishing the validity of the measurement instrument, we examined the structural model using LISREL. Figure 3 shows the standardised LISREL path coefficients and the overall fit indices. All the fit indices meet the recommended guidelines (Gefen et al., 2000). Thus, the structural model is an adequate fit for the data. In the usage model, usefulness and emotion were found to be insignificant to usage, while the perceived value and habit were significantly related to usage, explaining 59% of the usage variance. Therefore, H1a and H6a, but not H3a or H5a, were supported. In the continuance intention model, emotion was insignificant to continuance intention, while usefulness, perceived value and habit were significantly related to the continuance intention. These constructs explained 39% of the continuance intention variance. Consequently, H1b, H3b, H6b, but not H5b, were supported. In both models, usefulness and emotions were significantly related to the perceived value, so that H2 and H4 were supported.

Figure 3 Hypothesis testing: (a) usage model and (b) continuance intention model

(a)

422 H.W. Kim et al.

Figure 3 Hypothesis testing: (a) usage model and (b) continuance intention model (continued)

(b)

n.s.: not significant at 0.05 level. *p < 0.05; **p < 0.01 and ***p < 0.001.

6 Discussion

6.1 Discussion of findings

The relationship between emotion and behavioural intention has previously been tested and empirically supported by past studies (e.g., Kim et al., 2006). Therefore, the insignificance of the emotion-intention relationship is not due to its inexistence or inappropriateness, but rather to the context of the research and possibly gender bias (Dooley, 2001). In the collected data, only 32.4% of the respondents were female. Females score highest on feelings and emotions, and are more likely to spontaneously respond to their feelings than males (Majee and Hojat, 1998). From our results, the respondents were not prone to react to emotions, but they rather evaluated their emotions via the perceived value construct. Given that the majority of the respondents in our study were males, this result was in agreement with Majee and Hojat’s (1998) findings. The consumer behaviour literature demonstrates that what specifically determines the intention to consume depends on the utilitarian or hedonic nature of the product (Holt, 1995). Hedonic systems aim at providing self-fulfilling value to the users, while utilitarian systems aim at providing instrumental value to the users. Instrumentality implies there is an objective external to the interaction between user and system, such as increasing task performance and achieving expectations. Hedonic systems do not aim at facilitating such objectives. For utilitarian systems, the perceived usefulness is the strongest predictor of usage intention, at the expense of perceived emotion (Venkatesh and Davis, 2000). In the context of our research, mobile internet services may be considered more utilitarian than hedonic. In this case, the users will pay more attention to the usefulness of the service, and emotion will have less influence on intention.

In the usage model, usefulness-usage and emotion-usage relationships were both insignificant, indicating that the consumers did not directly react to usefulness and emotion, but they are more likely to evaluate them via the perceived value. This result is consistent with most studies in the same field, which do not support the direct relationship from cognition and affection to actual usage behaviour. However, in the


continuance model, usefulness has a direct significant effect on forming users’ continuance intention, indicating that consumers will consider usefulness factors when they intend to use the target system. In other words, usefulness has a strong predicting power on users’ continuance intention.

A comparison between the two models indicated that the perceived value has a stronger effect on continuance intention than usage, while habit has a stronger effect on usage than the continuance intention. The stronger predicting power of both usefulness and perceived value on continuance intention shows that the formation of continuance intention is a relatively more rational and conscious process. On the other hand, the stronger effect of habit on usage behaviour shows that the decision-making process of usage behaviour is less rational and rather unconscious, and, therefore, that mobile internet service users are probably using the services because they have already become accustomed to using them. However, when they are forming the usage intention, they consciously and more closely consider the benefits such as the overall evaluation of the service.

6.2 Limitations

As the implications of any study must be considered in the context of its limitations, we consider the study limitations before we discuss its implications. First, since emotional responses are not always recallable, answering emotions-related questions via a survey questionnaire may not yield accurate results. Furthermore, due to the time constraint, we were unable to conduct further quota sampling to ensure the absence of any gender bias in our sample (Dooley, 2001). Second, although we followed the age group distribution of mobile internet users from other countries, the sample may have been biased because most of the respondents were undergraduate students.

6.3 Implications

This research offers several theoretical and practical implications. According to Belk (1985) and Holbrook et al. (1990), behavioural models should be comprehensive and complete in order to be able to accurately predict behaviours in various situations. From the theoretical perspective, therefore, our framework extends past research by adopting rational, emotional and automatic components, including usefulness, emotion and habit, in predicting the continuance intention and usage behaviour. Many IS adoption and post-adoption studies even in the context of mobile services have largely focused on cognition-based behavioural models, such as TAM (Crabbe et al., 2009; Davis, 1989; Lin and Liu, 2009; Liu et al., 2008) and the IS continuance model (Bhattacherjee, 2001a; Hung et al., 2007; Lin and Shih, 2008). Those cognition-oriented models may play a critical role in explaining IS usage behaviour and intention in organisational settings where the cost of adoption and usage is handled by the organisation and most users are employees (Kim et al., 2007).

However, the emergence of the internet and mobile technologies has generated new forms of Information and Communication Technologies (ICTs) that are used in non-organisational settings. In that case, unlike users of traditional technologies, adopters of new ICT use the technology for personal purposes and the cost of adoption and usage is paid by the individuals. Such users play the dual role of technology users and service consumers, implying that both cognitive factors and affective factors should be

424 H.W. Kim et al.

considered to better account for usage behaviour and intention (Kim et al., 2007). Therefore, previous studies have advocated the extension of the traditional one-sided cognitive view by integrating the affective components of consumption experience (Holbrook et al., 1990; Kim et al., 2007).

While the integrated model reflecting both cognitive and affective views has been proposed for better prediction of the continuance intention and the usage behaviour, it has been argued that habitual factors should also be considered because a behaviour bypasses rational decision-making when it is repeatedly performed (Verplanken et al., 1998). As Triandis (1980) also found that the frequency of past behaviour is the best predictor of future behaviour, we suggested an extended model to include the habitual factor along with the cognitive and affective factors in the study of IS continuance intention and usage behaviour.

Furthermore, we used the same framework in order to analyse the difference between intention and usage behaviour. Psychological researches such as TAM, TPB and TRA mostly employed an intention-based model to predict behaviour, in which they assumed that behaviour is mostly predicted by intention. However, Belk (1985) insisted on a discrepancy between intention and actual behaviour. After testing some selected studies of intention-behaviour congruence, Belk found that the correlation between intention and behaviour in those studies was mostly below 0.5 and even much lower, which indicated the failure of intention in accurately predicting behaviour. The discrepancy between intention and actual behaviour occurs for a number of reasons, such as the possible change in individual preferences over time because new information becomes available. Therefore, we empirically compared intention and usage behaviour in this research within the same framework in order to examine the difference between them. Through this two-model comparison, a comprehensive framework was proposed as being more suitable in predicting usage behaviour than continuance intention (R2 of usage model = 0.59; R2 of continuance intention model = 0.39).

Our findings also have potential practical implications for mobile internet service providers. This research has provided some evidence concerning the cognitive, affective and habitual criteria that adopters use to evaluate whether to use or intend to use a specific IS. The research results showed that usefulness and emotion play important roles in accounting for both usage and intention through perceived value. Therefore, mobile internet service providers should focus on the instrumental consumption experiences and the emotional experiences of users to improve their perceived value. This implies that service providers need to create useful and enjoyable services to be updated constantly, thereby keeping customers in a loop of services. For the services to be useful, system design could focus on user needs and desires in ways that cannot be achieved via typical e-commerce (Kim et al., 2007). The unique characteristics of the particular technology, such as the time- and location-specific features of mobile internet services (Xu et al., 2009), could be used to achieve the provision of useful IS services as well.

In addition to enhancing the instrumental consumption experiences, since users favourably evaluated their emotions, contributing towards a higher perceived value of mobile internet services, service providers may have to offer services that elicit positive emotions such as pleasure and playfulness so that these feelings are used for their information value. Downloading ring tones, logos and games is classified as an active and hedonic service that generates a high level of positive emotions. There are several mobile services generating emotional experiences such as mobile video call services (Lin and Liu, 2009) and location-based services (Xu et al., 2009). Offering interactive


and multi-media interfaces and advertising the emotional aspects are examples of enhancing the users’ affective levels (Kim et al., 2007). Our findings also show that habit has a very significant effect on both actual usage and continuance intention. Encouraging frequent usage of mobile internet services until this usage becomes habitual is thus essential for customers to continue using the services. This can be achieved by providing free or cheap trials for a duration long enough to render the service usage habitual.

7 Conclusion

Mobile communication and services have widely been adopted and used worldwide. Research on mobile service usage has recently emerged as a practically important issue in IS literature. However, most previous studies on mobile internet services are cognition-oriented, while mobile internet services are characterised by both utilitarian and hedonic aspects. Usage experiences of mobile internet services incorporate diverse aspects such as cognitive, affective and judgemental because most users use mobile internet services for personal purposes and have to pay usage fees. Those diverse aspects would affect the way the users make decisions about the usage and continuance intention of mobile services. Few studies have examined the diverse aspects of the same research model.

This study has examined and compared mobile internet services usage and the continuance by developing an integrated model based on cognitive, affective and judgemental components, as well as automatic components such as habit. The study results revealed the determinants of mobile internet services usage behaviour and the continuance intention and how these determinants affect them. We have also examined the differences between the usage model and the continuance model by comparing them within the same model framework. The comparison of the two models indicated that the proposed model is more suitable for predicting usage behaviour than the continuance intention in the context of mobile internet services. By identifying the different influences of the determinants that drive usage behaviour and continuance intention, the study results have helped advance IS theory and offered insights for mobile internet service providers.

Acknowledgements

This work was supported by Yonsei University (2009-1-0140); the research programme 2010 of Kookmin University in Korea; and financially supported by Hansung University during 2010.

References Aarts, H., Verplanken, B. and Van Knippenberg, A. (1998) ‘Predicting behavior from actions in the

past: repeated decision making or a matter of habit?’, Journal of Applied Social Psychology, Vol. 28, No. 15, pp.1355–1374.

Ajzen, I. (1991) ‘The theory of planned behavior’, Organizational Behavior and Human Decision Processes, Vol. 50, No. 2, pp.179–211.

426 H.W. Kim et al.

Allen, C.T., Machleit, K.A. and Kleine, S.S. (1992) ‘A comparison of attitudes and emotions as predictors of behavior at diverse levels of behavioral experience’, Journal of Consumer Research, Vol. 18, No. 4, pp.493–504.

Bagozzi, R.P., Gopinath, M. and Nyer, P. (1999) ‘The role of emotions in marketing’, Academy of Marketing Science Journal, Vol. 27, No. 2, pp.184–206.

Belk, R.W. (1985) ‘Issues in the intention-behavior discrepancy’, Research in Consumer Behavior, Vol. 1, pp.1–34.

Bhattacherjee, A. (2001a) ‘An empirical analysis of the antecedents of electronic commerce service continuance’, Decision Support Systems, Vol. 32, pp.201–214.

Bhattacherjee, A. (2001b) ‘Understanding information systems continuance: an expectation-confirmation model’, MIS Quarterly, Vol. 25, No. 3, pp.351–370.

Cheung, W., Chang, M.K. and Lai, V.S. (2000) ‘Prediction of internet and world wide web usage at work: a test of an extended Triandis model’, Decision Support Systems, Vol. 30, pp.83–100.

Cohen, J.B. and Areni, C.S. (1991) ‘Affect and consumer behavior’, in Thomas, S.R. and Harold, H.K. (Eds.): Handbook of Consumer Behavior, Prentice-Hall, NJ, pp.188–240.

Crabbe, M., Standing, C., Standing, S. and Karjaluoto, H. (2009) ‘An adoption model for mobile banking in Ghana’, International Journal of Mobile Communications, Vol. 7, No. 5, pp.515–543.

Davis, F.D. (1989) ‘Perceived usefulness, perceived ease of use, and user acceptance of information technology’, MIS Quarterly, Vol. 13, No. 3, pp.319–340.

Derbaix, C. and Pham, M.T. (1991) ‘Affective reactions to consumption situations: a pilot investigation’, Journal of Economic Psychology, Vol. 12, pp.325–355.

Dooley, D. (2001) Social Research Methods, 4th ed., Upper Saddle River, Prentice-Hall, New Jersey.

Economides, A.A. and Grousopoulou, A. (2008) ‘Use of mobile phones by male and female Greek Students’, International Journal of Mobile Communications, Vol. 6, No. 6, pp.729–749.

Gardner, M.P. (1985) ‘Mood states and consumer behavior: a critical review’, Journal of Consumer Research, Vol. 12, pp.281–300.

Gefen, D., Straub, D.W. and Boudreau, M. (2000) ‘Structural equation modeling and regression: guidelines for research practice’, Communications of the Association for Information Systems, Vol. 4, No. 7, pp.2–76.

Holbrook, M.B., O’Shaughnessy, J. and Bell, S. (1990) ‘Action and reaction in then consumption experience: then complementary role of reasons and emotions in consumer behavior’, Research in Consumer Behavior, Vol. 1, pp.1–34.

Holt, D.B. (1995) ‘How consumers consume: a typology of consumption practices’, Journal of Consumer Research, Vol. 22, No. 1, pp.1–16.

Hung, M., Hwang, H. and Hsiej, T. (2007) ‘An exploratory study on the continuance of mobile commerce: an extended expectation-confirmation model of information system use’, International Journal of Mobile Communications, Vol. 57, No. 4, pp.409–422.

Husig, S. and Hipp, C. (2009a) ‘The disruptive potential of Wi-Fi in the German hotspot market’, International Journal of Mobile Communications, Vol. 6, No. 6, pp.729–749.

Husig, S. and Hipp, C. (2009b) ‘The disruptive potential of Wi-Fi in the German hotspot market’, International Journal of Mobile Communications, Vol. 7, No. 6, pp.615–631.

Kang, J., Ko, Y. and Ko, I. (2009) ‘The impacts of social support of guild members and psychological factors on flow and loyalty in MMORPG’, Asia Pacific Journal of Information Systems, Vol. 19, No. 3, pp.45–76.

Karahanna, E., Straub, D.W. and Chervany, N.L. (1999) ‘Information technology adoption across time: a cross-sectional comparison of pre-adoption and post-adoption beliefs’, MIS Quarterly, Vol. 23, No. 2, pp.183–213.


Kim, H.W., Chan, H.C. and Chan, Y.P. (2007) ‘A balanced thinking-feelings model of Information systems continuance’, International Journal of Human-Computer Studies, Vol. 65, No. 6, pp.511–525.

Kim, H.W., Chan, H.C. and Gupta, S. (2006) ‘Value-based adoption of mobile internet: an empirical investigation’, Decision Support Systems, Vol. 43, No. 1, pp.1–16.

Kivi, A. (2009) ‘Measuring mobile service usage: methods and measurement points’, International Journal of Mobile Communications, Vol. 7, No. 4, pp.3415–435.

Lazarus, R.S. (1991) Emotion and Adaptation, Oxford University Press, New York. Lee, H.Y. and Ahn, H. (2009) ‘A study on EC acceptance’, Asia Pacific Journal of Information

Systems, Vol. 19, No. 1, pp.147–180. Lee, T. and Jun, J. (2007) ‘The role of contextual marketing offer in mobile commerce acceptance:

comparison between mobile commerce users and nonusers’, International Journal of Mobile Communications, Vol. 5, No. 3, pp.339–356.

Li, W. and McQueen, R. (2008) ‘Barriers to mobile commerce adoption: an analysis framework for a country-level perspective’, International Journal of Mobile Communications, Vol. 6, No. 2, pp.231–257.

Liao, C., Chen, J-L. and Yen, D.C. (2007) ‘Theory of Planning Behavior (TPB) and customer satisfaction in the continued use of e-service: an integrated model’, Computers in Human Behavior, Vol. 23, No. 6, pp.2804–2822.

Limayem, M. and Hirt, S.G. (2003) ‘Force of habit and information systems usage: theory and initial validation’, Journal of the Association for Information Systems, Vol. 4, pp.65–97.

Limayem, M., Hirt, S.G. and Cheung, C.M.K. (2003) ‘Habit in the context of IS continuance: theory extension and scale development’, Paper presented in Proceedings of the 11th European Conference on Information Systems, Naples, Italy, pp.90–108.

Limayem, M., Hirt, S.G. and Cheung, C.M.K. (2007) ‘How habit limits the predictive power of intention: the case of information systems continuance’, MIS Quarterly, Vol. 31, No. 4, pp.705–737.

Lin, J. and Liu, E. (2009) ‘The adoption behavior for mobile video call services’, International Journal of Mobile Communications, Vol. 7, No. 6, pp.646–666.

Lin, Y. and Shih, D. (2008) ‘Deconstructing mobile commerce service with continuance intention’, International Journal of Mobile Communications, Vol. 6, No. 1, pp.67–87.

Liu, M., Huang, L. and Chen, A. (2008) ‘Chinese consumer’s adoption intention towards 3G mobile phone’, International Journal of Mobile Communications, Vol. 6, No. 6, pp.770–786.

Majee, M. and Hojat, M. (1998) ‘Personality profiles of male and female positive role models in medicine’, Psychology Reports, PMID: 9821729 (Pub Med-indexed for MEDLINE), Vol. 82, No. 2, pp.547–559.

Nickerson, R., Isaac, H. and Mak, B. (2008) ‘A multi-national study of attitude about mobile phone use in social setting’, International Journal of Mobile Communications, Vol. 6, No. 6, pp.541–563.

Ouellette, J.A. and Wood, W. (1998) ‘Habit and intention in everyday life: the multiple processes by which past behavior predicts future behavior’, Psychological Bulletin, Vol. 124, No. 1, pp.54–74.

Palmgreen, P. (1984) ‘Uses and gratifications: a theoretical perspective’, in Bostrom, R.N. (Ed.): Communication Yearbook, Sage Publications, Beverly Hills, CA, Vol. 8, pp.61–72.

Parveen, F., Abessi, M. and Ainin, S. (2009) ‘Wireless internet-using mobile devices in Malaysia’, International Journal of Mobile Communications, Vol. 7, No. 5, pp.580–593.

Romer, P.M. (2000) ‘Thinking and feeling’, The American Economic Review, Vol. 90, No. 2, pp.439–443.

Safko, L. and Brake, D.K. (2009) The Social Media Bible: Tactical, Tools and Strategies for Business Success, Wiley, New Jersey.

428 H.W. Kim et al.

Schwarz, N. (2000) ‘Attitude and social judgment’, European Journal of Social Psychology, Vol. 30, pp.149–176.

Sheeran, P. and Orbell, S. (1999) ‘Implementation and intentions and repeated behavior: augmenting the predictive validity of the theory of planned behavior’, European Journal of Social Physiology, Vol. 29, pp.349–369.

Sirdeshmukh, D., Singh, J. and Sabol, B. (2002) ‘Consumer trust, value, and loyalty in relational exchange’, Journal of Marketing, Vol. 66, No. 1, pp.15–37.

Thaler, R. (1985) ‘Mental accounting and consumer choice’, Marketing Science, Vol. 4, No. 3, pp.199–214.

Triandis, H.C. (1980) ‘Values, attitudes, and interpersonal behavior’, in Howe Jr., H.E. and Page, M.M. (Eds.): Beliefs, Attitudes and Values, Nebraska Symposium on Motivation, 1979, 27, University of Nebraska Press, Lincoln, pp.195–259.

Venkatesh, V. and Davis, F.D. (2000) ‘A theoretical extension of the technology acceptance model: four longitudinal field studies’, Management Science, Vol. 46, No. 2, pp.186–204.

Verplanken, B., Aarts, H., van Knippenberg, A. and Moonen, A. (1998) ‘Habit versus planned behavior: a field experiment’, British Journal of Social Psychology, Vol. 37, pp.111–128.

Xu, Z. and Yuan, Y. (2009) ‘The impact of context and incentives on mobile service adoption’, International Journal of Mobile Communications, Vol. 7, No. 3, pp.363–381.

Xu, H., Oh, L. and Teo, T. (2009) ‘Perceived effectiveness of text vs. multimedia location-based advertising messaging’, International Journal of Mobile Communications, Vol. 7, No. 2, pp.154–177.

Zajonc, R.B. (1980) ‘Feeling and thinking: preferences need no inferences’, American Psychologist, Vol. 35, No. 2, pp.151–175.

Zeithaml, V.A. (1988) ‘Consumer perceptions of price, quality and value: a means-end model and synthesis of evidence’, Journal of Marketing, Vol. 52, No. 3, pp.2–22.

Appendix: Survey measurement items

Variable Item Description Reference

INT1 I intend to continue using mobile internet services in the next six months

INT2 I expected my use of mobile internet services to continue in the future

INT3 During the next six months, I plan to continue using mobile internet services

Continuance intention

INT4 If I could, then I would like to discontinue my use of mobile internet services

Bhattacherjee (2001b)

How do you feel about the use of mobile internet services? EMO1 Unsatisfied – Satisfied EMO2 Annoyed – Pleased EMO3 Frustrated – Contented EMO4 Unhappy – Happy

Emotion

EMO5 Bored – Excited

Cohen and Areni (1991) and Cheung et al. (2000)

HAB1 The use of mobile internet services has become a habit for me

Habit

HAB2 I am addicted to using mobile internet services


Appendix: Survey measurement items (continued)

Variable Item Description Reference

HAB3 I must use mobile internet services HAB4 I do not even think twice before using mobile

internet services

Habit

HAB5 Using mobile internet services has become natural to me

Limayem and Hirt (2003)

VAL1 Considering the fee I pay, the use of mobile internet services offers value for money

VAL2 Considering the time and effort I spend, the use of mobile internet services is worthwhile to me

VAL3 Considering all monetary and non-monetary costs, the use of mobile internet services is of good value

Perceived value

VAL4 Overall, the use of mobile internet services delivers me good value

Sirdeshmukh et al. (2002)

USF1 Using mobile internet services enables me to complete my tasks

USF2 Using mobile internet services makes it easier for me to complete my tasks

USF3 Using mobile internet services saves me time and effort in performing tasks

Usefulness

USF4 Mobile internet services are useful to me in performing my task

Davis (1989)

USG1 I use mobile internet services very frequently (many times per month)

USG2 I use mobile internet services for a variety of tasks USG3 I use mobile internet services very intensively

(many minutes per month)

Usage

USG4 Overall, I use mobile internet services a lot

Cheung et al. (2000)

Documents

An integrated model of mobile internet services usage and