Study of How Experience Involvement Affects Users’ Continuance Intention to Use Mobile Reading

1 Introduction

Along with rapid development of computer technology and network technology, digital information products have increasingly become the mainstream goods. Mobile reading is a typical example of digital information products. This paper will take mobile reading as an example to study the impacts of consumer involvement upon their continuous intention. Mobile reading is based on mobile phones (WAP, APP), tablets and E-books. Before reading, consumers can gratuitously experience 10∼20 chapters, and such continuous experience is called as consumer involvement. Different involvements are highly related to purchase behavior, reflecting the higher one person’s self-involvement of one thing and the higher possibility for acceptance of such products.

Today, due to rapid development of mobile Internet technologies and user’s multiple choices, it is a difficult issue for sustained business development on how to retain customers and keep them to permanently use mobile reading. So it is very urgent to study the factors affecting mobile reading effect according to user’s needs and reading contents.

2 Literature review

Fishbein and Ajzen[5] proposed the theory of reasoned action (TRA) to predict whether human beings would be engaged in some actions. According to TRA, human’s behavioral intention (BI) decides human’s behavioral expression, and individual’s attitude toward such behavior and subjective norms (SN) reflecting social impact can affect the BI and attitude. The more positive attitudes and subjective norms are, the higher the BI is, or vice versa[5].

Technology acceptance model (TAM) is a theoretical model that Davis[3] used TRA to study user’s acceptance of information system, and it is an amendment of TRA. TAM introduces “perceived usefulness” and “perceived ease of use” to explain user’s attitude and intention. Davis et al[18] also introduced perceived enjoyment into TAM. In addition, it is discovered that perceived ease of use can, through the regulating roles of perceived enjoyment, affect the BI. After adding the perceived enjoyment variable, the model’s explanatory power has been improved significantly.

Expectation confirmation theory (ECT) was proposed by Oliver[12] on the basis of cognitive dissonance theory (CDT). According to ECT, customer’s repeated purchase intentions are closely related to their past experience, and satisfactory experience is the key for them to establish long-term relationships with products or service suppliers. ECT is a basic theory to study consumer’s satisfaction, as it elaborates the confirmation between expectation and perceived performance, and then judges whether customers are satisfied with products or services. Whereas, the satisfaction will be a reference for next purchase or use (repurchase intention). ECT has been widely used in the studies on customer satisfaction and repurchases intention. The expectation confirmation model (ECM) is shown in Figure 1.

Figure 1

Expectation confirmation model

To make ECT more applicable to information system environment, Bhattacherjee[2] also drew on Davis’ TAM and then modified ECT, thus got ECM-IT for continuous use of information system, which is shown in Figure 2.

Figure 2

ECM-IT model

3 Research model and assumptions

3.1 Research model

According to theoretical overview, we can expand and amend specific connotation of expectation/belief variable in ECM-IT model, broaden the scope of research and improve the degree of explanation of the model, thus better predict and explain user’s intention. In previous studies, some scholars indicated that consumer’s expectations included the properties of products or services themselves, efforts made for the gains, costs and social benefits , so the connotation of expectation and user’s perceived value was extremely similar. In addition, only with perceived usefulness and perceived ease of use as user’s perception of mobile reading cannot reflect main features of mobile reading. As mobile reading is a content-based business, and its value is that the contents bring perceived values to users, but such value is undefined. Thus, the in-depth understanding of user’s perceived value is helpful for enterprisers to carry out accurate value orientation and then better create values, playing dominated roles in the entire value network.

In this paper, user’s perceived value based on mobile reading is considered as specific contents of expectation/belief variable after ECM-IT is used, namely, six dimensions of user’s perceived value based on mobile reading are used to replace perceived usefulness in the ECM-IT model. Meanwhile, this paper amends the ECM-IT model according to the characteristics of mobile reading, and also adds subjective norm factors in the model; furthermore, as user’s satisfaction is constantly changing during mobile reading experience process, we cannot determine user’s satisfaction after experience. Therefore, the variable of satisfaction is removed in this paper. The conceptual model in this study is shown in Figure 3.

Figure 3

Research model

4 Research methodology

In this paper, we use Likert7 to measure variables. Respondents are required to select strongly disagree, disagree, somewhat disagree, not sure, somewhat agree, agree and strongly agree according to their real mobile reading conditions, and then give 1∼7 scores respectively. The latent and observed variables are shown in Table 1. For the objectivity of research, the investigated objects are identified as old users or active users. In consideration of youth’s needs and interests on mobile reading business, this study mainly focuses upon 18∼35 years old people with particular consumption power, coupled with a certain amount of users at other ages. We distributed 240 questionnaires totally, and recovered 181 effective questionnaires, thus the effective recovery rate is 75%. Among the samples, there are 103 men, accounting for 56.91%; and 78 women, accounting for 43.09%. So the proportion between men and women is coordinated. This paper mainly uses SPSS19.0 analysis software to make preliminary analysis of surveyed data, and meanwhile adopts factor analysis and reliability analysis to verify the reliability and validity of data. Subsequently, we use AMOS18.0 structural equation approach to verify the data fitting and conceptual model, and also carry out relationship analysis between variables according to research results.

5 Reliability and validity analysis

Reliability means the degree of stability and consistency of measured results. The higher the reliability, the lower the impact of measured scores due to errors. Generally, it is reliable if Cronbach’s Alpha coefficient is greater than 0.6. Our test results show that Cronbach’s Alpha coefficient of each variable is greater than 0.6, indicating that measuring indictors of latent variables has reliable internal consistency.

Validity means the validity of test results, that is, researchers need to know whether the questionnaires are tested effectively. This study uses factor analysis method to verify the scale structural validity. Factor analysis needs to meet the following two conditions: Bartlett’s Test of Sphericity and KMO (Kaiser-Meyer-Olkin) value identification.

After the above two tests, this study uses “principal component factor analysis” approach, and takes Varimax as a rotation way, eigenvalue (>1) as the standard to choose the number of factors, and variable’s factor loading (>0.5) as the standard for factor analysis. This study will carry out independent factor analysis of questions relating to dependent variables in conceptual model, and also carry out factor analysis of questions relating to precursor variables in dependent variables. Table 2 shows the results of validity analysis.

From principal component factor analysis, we extract NINE common factors with eigenvalue greater than one. Eigenvalue of common factor 1 is 3.906, descending from 1 to 8, and the eigenvalue of common factor 9 is 4.561. The accumulatively explanatory population variance of these nine factors is 83.145%, as shown in Table 3.

The factor loading on associated variables is greater than 0.5, with good convergent validity, showing good discriminant validity.

6 Result analysis

This paper uses structural equation to process the model, calculate standardized regression coefficients (path coefficient) and significance of each path, and builds the model’s path relationship diagram, thus determine causal relationship between independent variables, intermediate variables and dependent variables. There are nine latent variables in the model, so we can use these 27 observable variables to indirectly measure these nine latent variables. AMOS’ operating results give the model’s commonly used fitting indicators, shown in Table 4.

According to model indicators, the fitting degree of mobile reading user’s continuous intention model is good, and all evaluation indicators are within the acceptable range. Therefore, such model is acceptable statistically.

This paper uses path analysis model of AMOS19.0 structural equation — recursive model assumptions for confirmatory analysis. Due to the assumption of recursive model, there is traceability relationship between variables. The result is in Table 5.

It can be seen from the above table, the model’s overall validation effect is better, and among 14 assumptions in the model, the significance of 10 assumptions has been verified.

From significance relationship of the assumptions, except that the influence significance of perceived ease of use on continuous intention, influence significance of perceived risk on continuous intention, influence significance of experience involvement on perceived image enhancement, and influence significance of experience involvement on perceived ease of use are larger than 0.05 (p > 0.05), the significance of other relationships is less than 0.05 (p < 0.05).

In general, the model’s overall validation effect is good. After removing non-significant assumptions, we can get the modified model as shown in Figure 4.

Figure 4

Significant structural variable chart

According to the results of AMOS19.0, we verified our hypotheses, and found that except that H3c, H4b, H6b, H6c are false, other assumptions are true.

There are four non-significant relationships. First, Influence of experience involvement on perceived ease of use is not significant, because mobile reading is pretty convenient to download and operate, thus experience won’t have great effects on consumers. Besides, for mobile reading, contents is the king. Operation and interface is less important than the usefulness and pleasure of its contents for mobile reading users. Therefore, the influence of perceived ease of use on continuous intention is not significant. And the influence of experience involvement on perceived ease of use is also not significant, the possible reason is that users’ perception is changing all the time in the process of experience reading. As for the non-significant influence of perceived risk on continuous intention, the paper argues that at the present stage users mainly focus on the contents rather than risks like privacy and unknown charge. But with the further development of mobile reading, risks must be the main factors influencing users.

7 Result analysis

For mobile reading, reading several chapters for free is necessary for consumers to know if the content is worthy for them.

In this paper, we study the influential factors for continuous use of mobile reading business from the perspective of experience involvement. We build the variables from the perceived values and perceived efforts, establish the appropriate structural model based on extended ECM-IT theory, and then come to the following conclusions through verification: 1) Subjective norm significantly affect experience involvement, whereas experience involvement will significantly and directly affect user’s continuous intention, and meanwhile indirectly affect user’s continuous intention by affecting the perceived value of mobile reading business; 2) Experience involvement significantly affect user’s perceived usefulness, perceived enjoyment, perceived cost level and perceived risks; 3) User’s perceived usefulness, perceived enjoyment, perceived image enhancement and perceived cost level can significantly affect users’ continuous intention for mobile reading business.

According to the above conclusions, we suggest the mobile reading providers pay more attention to the free experience process, increasing its convenience, enriching its content and rationalizing its cost. In this way, mobile reading providers can promote users’ perceived usefulness, perceived enjoyment and perceived cost level which can affect users’ continuous intention directly and significantly. But according to our findings, experience involvement can’t significantly affect users’ perceived image enhancement. Possibly because the current mobile reading can’t give users the perception of image enhancement. Users’ perceived image enhancement is very important under Maslow’s need hierarchy theory, because image enhancement belongs to self-actualization, which is the highest need hierarchy of Maslow. So mobile reading providers should optimize their mobile reading to improve users’ perceived image enhancement. We are now in the age of web2.0, users’ sense of participation is an important sign. Mobile reading providers should let users more involved to achieve value co-creation. When users have read several chapters, they can set the rest framework of the novel, and other users can choose to agree or not. The author will refer to the users’ vote, once the author adopted one of the users’ advice, the user will act as a co-author. Thus, users’ perceived image enhancement can be highly improved through their free experience process.

Our study is not only suitable for mobile reading, can also be applied to other areas, such as mobile phone manufacturing industry, e-commerce, etc. As mentioned above, sense of participation is very important, value co-creation is pretty hot, so our further study will focus on this field.