Integrating perceived playfulness into expectation-confirmation model for web portal context PDF

Preview

Summary

Information & Management 42 (2005) 683–693 www.elsevier.com/locate/dsw Integrating perceived playfulness into expectation-confirmation model for web portal context Cathy S. Lina, Sheng Wub, Ray J. Tsaic,* a National University of Kaohsiung, Kaohsiung 811, Taiwan b National Sun Yat-Sen University...

Description

Accelerat ing t he world's research.

Integrating perceived playfulness into expectation-confirmation model for web portal context Cathy Lin Information & Management

Cite this paper

Downloaded from Academia.edu 

Get the citation in MLA, APA, or Chicago styles

Related papers

Download a PDF Pack of t he best relat ed papers 

Cont inuance int ent ion of blog users: t he impact of perceived enjoyment , habit , user involvem… Wen-Lung Shiau Underst anding e-learning cont inuance int ent ion: An ext ension of t he Technology Accept ance Model Juan Roca driving-fact ors-for online shopping among-st udent in Saudi Arabia Talal Almaghrabi

Information & Management 42 (2005) 683–693 www.elsevier.com/locate/dsw

Integrating perceived playfulness into expectation-confirmation model for web portal context Cathy S. Lina, Sheng Wub, Ray J. Tsaic,* a

National University of Kaohsiung, Kaohsiung 811, Taiwan National Sun Yat-Sen University, Kaohsiung 804, Taiwan c BCIS Department, St. Cloud State University (SCSU), 720, 4th Ave. S., St. Cloud, MN 56301, USA b

Received 1 December 2002; received in revised form 23 April 2003; accepted 25 April 2004 Available online 2 September 2004

Abstract This paper investigated the value of including ‘‘playfulness’’ in expectation-confirmation theory (ECT) when studying continued use of a web site. Original models examined cognitive beliefs and effects that influence a person’s intention to continue to use an information system. Here, an extended ECT model (with an additional relationship between perceived playfulness and satisfaction) was shown to provide a better fit than a simple path from perceived usefulness to satisfaction. The results indicated that perceived playfulness, confirmation to satisfaction, and perceived usefulness all contributed significantly to the users’ intent to reuse a web site. Thus, we believe that the extended ECT model is an appropriate tool for the study of web site effects. # 2004 Published by Elsevier B.V. Keywords: Web portal; Expectation-confirmation theory; Perceived playfulness; Satisfaction

1. Introduction Today, the World Wide Web (WWW) impacts many facets of our lives, including communication, entertainment, social activities, shopping, etc. The web portal is the most accessed type of site and is advertising-supported [51]; the more users who visit the site, the more income it generates [50]. To remain competitive, therefore, web portal providers have to be creative and innovative in attracting users. Companies * Corresponding author. Tel.: +1 320 255 4169; fax: +1 320 308 6074. E-mail address: [email protected] (R.J. Tsai). 0378-7206/$ – see front matter # 2004 Published by Elsevier B.V. doi:10.1016/j.im.2004.04.003

provide special services, such as free email, personal web space, searches, content provision, etc. and identify users’ tastes, needs, and other purchasing behaviors to hold their attention. Obtaining and maintaining user loyalty is the key to survival. Today user perception to a web site is very important; much research has focused on the Internet users’ behavior. Users have many options that potentially reduce their loyalty; they can go anywhere and may not return. Thus, retaining users becomes both important and difficult. Some well-known theories, such as the technology acceptance model (TAM) [15] and the theory of planned behavior [2] have been used to examine

684

C.S. Lin et al. / Information & Management 42 (2005) 683–693

variables that motivate individuals to accept and use an IS. However, the characteristics of portal surfing are different from business use of an IS; the Internet surfers are, to some extent, irregular in their actions and may not return once they leave [10]. Furthermore, in the web context, attracting users and maintain their loyalty is critical. To study continued use, expectationconfirmation theory (ECT) can be used to explore the post-behavior aspect of a sale. Specifically, once a user has positive feelings about a web site, the satisfaction results in a revisit intention. Although previous studies [8,33,38] suggested that perceived usefulness plays an important role in predicting users intentions, it is not known whether this is a dominant factor in web surfing for the general public. Since thousands of portals provide similar services, retention of existing users is difficult. Attracting and retaining users by providing a playful web site has gained researchers’ attention. Playfulness is significantly associated with total web use, especially for entertainment purposes [3]. Moon and Kim [41] extend TAM for a WWW context and indicated that playfulness was an intrinsic motivation factor in acceptance of the site. Therefore, we integrated the ‘‘perceived playfulness’’ construct into ECT in an attempt to enhance understanding of individuals’ web portal revisit intentions.

between actual performance and expectations was not perceived as being small, the perceived performance would increase or decrease with expectation (as indicated by the arrow from expectations to perceived performance). Third, the perceived performance may therefore either confirm or refute pre-purchase expectations. Fourth, as shown in Fig. 1, expectation and the perceived level of confirmation positively affects satisfaction. If confirmation occurred, customers’ satisfaction increased or decreased from this baseline level. Finally, satisfied consumers formed an intention to repurchase while dissatisfied users did not. Applying the ECT framework to the study of users’ behaviors in an e-commerce context is appropriate [14,47,55] because users’ behavior affects their repurchase decision; however, some theoretical extensions may be required to explain users’ continuance behavior in using a web site. Therefore, the model here focused only on post-acceptance variables, since the expost expectations are especially important for web portals, because users’ expectations may change over time. If a web portal targets users and heightens their loyalty to the site, it is necessary to understand the users’ continuance behavior. Hence, the ex-post expectation is represented by perceived usefulness, which is a cognitive belief salient to IS use and consistent with ECT’s definition of expectation. As in attitude theories, the (ex-post) expectation represents the beliefs weighted with evaluation of outcomes [46].

2. Theoretical background 2.2. Playfulness 2.1. Expectation-confirmation theory ECT has been widely used to study consumer satisfaction and post-purchase behavior (e.g., repurchase, complaining), service marketing in general [1,39,43,44,48], and online banking users’ reuse of an information system [6]. In sum, it was used to examine both pre-behavior (expectation) and postbehavior (perceived performance) variables rather than only pre-behavior. Fig. 1 illustrates key constructs and relationships of ECT. Oliver [45] described the process by which consumers reached repurchase intentions. First, consumers formed (ex-ante) expectations of the product or service prior to purchasing it. Second, the purchase resulted in an (ex-post) perceived performance that was influenced by these expectations. If the difference

Characterizing playfulness is difficult [5], because laymen and researchers use the term ‘‘play’’ in several ways [17,20]. The trait of playfulness may be treated as a motivational characteristic. Individuals who ranked high in playfulness demonstrate better performance and higher affective response to computer training tasks. But playfulness may instead be defined as a situational characteristic of the interaction between an individual and a situation. The majority of research on it as an individual’s interaction state is based on Csikszentmihalyi’s [13] flow theory. He notes that ‘‘people seek flow experience primarily for itself;’’ thus, flow serves as a positive reinforcer that increases the probability of future usage in a computer-mediated environment (CME) [23,27]. The theory has also been proposed as a way to understand consumer navigation behavior

C.S. Lin et al. / Information & Management 42 (2005) 683–693

685

Fig. 1. Expectation-confirmation theory. Note: t1 = pre-consumption variable; t2 = post-consumption variable.

in online environments. Here, playfulness was regarded as an individual state, because an individual can feel more or less playful at various points during his/her visit to a web portal.

3. Research model and hypotheses This research model (Fig. 2) was therefore an extension of the original ECT based on individuals’ intrinsic motivation theory. Perceived playfulness, the extended part of the model, was the construct of interest. When people get involved in an activity for pleasure and enjoyment, this is their intrinsic motive and it should increase the likelihood that they would revisit the web portal in the future. Therefore, one basic assumption was that playfulness impacted individuals’ satisfaction in using web portals and thus their intention to reuse them. Deci and Ryan [18] suggested that people are both intrinsically and extrinsically motivated. Perceived

playfulness is an intrinsic motive, whereas perceived usefulness is extrinsic. Based on the original ECT model, perceived usefulness is influenced by confirmation. Since perceived usefulness and perceived playfulness are common motivations, it is reasonable to believe that confirmation would impact on perceived playfulness. Theoretical support for this association comes from cognitive dissonance theory [21], which suggests that users may experience cognitive dissonance or psychological tension if their pre-acceptance playfulness perceptions are not confirmed during actual use. Rational users may try to reduce this dissonance by distorting or modifying their playfulness perceptions to be more consistent with reality. Hence, this leads to the first hypothesis. H1. Users’ extent of confirmation is positively associated with their perceived playfulness of a web portal. Researchers argue that people who rank higher in playfulness will interact more playfully; they exercise and develop skills through exploratory behaviors,

Fig. 2. Research model.

686

C.S. Lin et al. / Information & Management 42 (2005) 683–693

resulting in enhanced task performance [40]. Previous research has shown that higher degrees of pleasure and involvement during computer interactions lead to concurrent subjective perceptions of positive affects and satisfaction. Webster et al. [52,55] noted that ‘‘higher playfulness results in immediate subjective experiences such as positive mood and satisfaction.’’ Playfulness is also positively associated with positive affect and satisfaction [34]. Sandelands et al. [49] found that attitudinal outcomes, such as positive affect, pleasure, and satisfaction, resulted from playful experiences. This leads to the second hypothesis. H2. Users’ perceived playfulness is positively associated with their satisfaction with a web portal. According to flow theory, a positive subjective experience is an important reason for performing an activity. If an individual ‘‘feels good’’ about an activity, it is intrinsically motivating, and he/she is more likely to engage in it. Individuals using a web portal and experiencing playfulness are more absorbed and interested in the interaction. This shapes their intention to visit this portal again later. Davis et al. [15,16] found that perceived playfulness explained significant variance in usage intentions. Corresponding to Igbaria’s study [29], Webster et al.’s study showed that playfulness was highly correlated with voluntary use. Similar findings revealed that perceived playfulness was positively related to microcomputer usage. This leads to our third hypothesis. H3. Users’ perceived playfulness of web portal use is positively associated with their continuance intention. Since ECT is used as a baseline model, the research verified the ECT hypothesized relationships in the context of web portals. In his empirical study of EC service continuance, Bhattacherjee employed the ECT model to study customer satisfaction and continuance. His results suggested that it was feasible to apply ECT to the web context. In addition, studies [25,53] on post-adoption behavior in online services showed the appropriateness of the study. Some related hypotheses were therefore formulated based on the baseline model of ECT (Fig. 2). H4. Users’ extent of confirmation is positively associated with their perceived usefulness of a web portal.

H5. Users’ extent of confirmation is positively associated with their satisfaction with a web portal. H6. Users’ perceived usefulness of a portal is positively associated with their satisfaction with the web portal. H7. Users’ perceived usefulness of a portal use is positively associated with their portal continuance intention. H8. Users’ satisfaction with the initial web portal is positively associated with their continuance intention.

4. Research methodology 4.1. Subjects Subjects of this research were individuals who used the web portal in their free time. Questionnaires were administered to senior level undergraduate students from three universities. The sample consisted of 300 undergraduate students majoring in management. All subjects had prior experience with web portal use. The survey indicated that students’ responses would be kept confidential and only summary information would be presented; administrators reiterated this verbally. Incomplete questionnaires were discarded, leaving 254 usable samples. Table 1 details the respondents’ characteristics. 4.2. Measurement development Five constructs were measured: perceived usefulness, confirmation, perceived playfulness, satisfaction, and IS continuance intention. Constructs were measured via a multiple-item scale drawn from pre-validated IS or ECT measures, and reworded to relate specifically to web portal use. The applicability of the modified items was enhanced by literature reviews, using three PhD students as expert judges, and pilot testing relevant samples. During this process, scale items were trimmed and refined, and dimensions were modified to hold the content validity as understanding of the constructs improved. All items used seven-point Likert scales anchored from ‘‘strongly disagree (=1)’’ to ‘‘strongly agree (=7)’’. Tables 2 and 3 provide operational definitions and scale items for these constructs.

687

C.S. Lin et al. / Information & Management 42 (2005) 683–693 Table 1 Descriptive statistics of respondents’ characteristics (N = 254) Measure

Items

Frequency

Percent

Gender

Female Male

157 97

61.8 38.2

Years of Internet experience

5 years

46 123 71 14

18.1 48.4 28 5.5

Primary web portal used

Yahoo Taiwan (tw.yahoo.com) Sina (www.sina.com.tw) Pchome (www.pchome.com.tw) Yam (www.yam.com.tw) Other

229 6 5 12 2

90.2 2.4 2 4.7 0.8

Time of use at primary web portal (daily)

9 h

146 99 7 1 1

57.5 39 2.8 0.4 0.4

Table 2 Operationalization of constructs Construct

Operational definition

Source

Perceived usefulness Confirmation

Users’ perception of the expected benefits of web portal use Users’ perception of the congruence between expectation of web portal use and its actual performance The strength of one’s belief that interacting with the web portal will fulfill the user’s intrinsic motives Users’ affect with (feelings about) prior web portal use Users’ intention to continue using web portal

Adapted from Davis et al. [15] Adapted from Bhattacherjee [6]

Perceived playfulness Satisfaction Continuance intention

Adapted from Moon and Kim [41] Adapted from Spreng and Olshavsky [50] Adapted from Bhattacherjee [6]

Table 3 Summary of measurement items Construct

Measurement items

Source

Perceived usefulness

Using web portal improves performance of my learning Using web portal improves skill of my search data Using web portal enables me to access a lot of usefulness information

Adapted from Davis et al. [15]

Confirmation

My experience with using the web portal was better than what I expected The service level provided by the web portal was better than what I expected Overall, most of my expectations from using the web portal were confirmed

Adapted from Bhattacherjee [6]

Perceived playfulness

When interacting with the web portal, I am not aware of the time as it elapses When interacting with the web portal, I am not aware of distracting noise When interacting with the web portal, I often forget other commitments

Adapted from Moon and Kim [41]

Satisfaction

Using Using Using Using

Adapted from Spreng and Olshavsky [50]

Continuance intention

I intend to continue using the web portal rather than discontinue its use My intentions are to continue using the web portal rather than use any alternative means

the the the the

web web web web

portal portal portal portal

makes makes makes makes

me me me me

feel feel feel feel

very very very very

satisfied pleased contented delighted

Adapted from Bhattacherjee [6]

688

C.S. Lin et al. / Information & Management 42 (2005) 683–693

5. Data analysis and results To assess the hypotheses, our research relied extensively on confirmatory factor analysis (CFA) using LISREL 8.30 and sample correlation matrix [31]. The following common measures were used to assess the model’s overall goodness of fit: chi-square/degree of freedom, goodness-of-fit index (GFI), normed fit index (NFI), comparative fit index (CFI), and standardized root mean square residual (SRMSR) [4]. Using these measures, we assessed the measurement model and determined whether the measured variables reliably reflected the theoretical constructs. Further, we checked the proposed research model’s overall goodness-of-fit compared to the original ECT model. 5.1. Assessment of measurement model The model was further assessed for construct reliability and validity. The resultant coefficient is similar to Cronbach’s a, except that it used the actual factor loadings rather than assuming each item was equally weighted in the composite load determination. As shown in Table 4, construct reliability for all the factors in our measurement model was above 0.7, an acceptable threshold suggested by Nunnally and Bernstein [42]. Construct validity for the five measurement scales (portal continuance intention, satisfaction, perceived usefulness, perceived playfulness, and confirmation) was assessed via CFA. Each scale item was modeled as a reflective indicator of its hypothesized latent construct. The five constructs were allowed to covary Table 4 Scale properties and correlations Construct Number of Reliabilitya AVEb Factor correlations items PU PP PCI S PU PP PCI S C

3 3 2 4 3

0.84 0.86 0.84 0.93 0.78

0.65 0.67 0.72 0.77 0.55

1 0.63 0.51 0.53 0.51

C

1 0.54 1 0.62 0.58 1 0.55 0.44 0.60 1

PU: perceived usefulness; PP: perceived playfulness; PCI: portal continuance intention; S: satisfaction; C: confirmation. P a 2 P = ( standardized loading) / P Composite reliability 2 ( standardized loading) + ej. P b Average variance extracted (AVE) = (standardized loading2)/ P P 2 (standardized loading) + ej.

in the CFA model. Model estimation used the maximum likelihood approach with the item correlation matrix used as input. The first s...