Influence of Users’ Technology Readiness on the Perceived Value of Mobile Phone Enabled Internet Services

The objective of this study was to investigate the influence of users’ technology readiness (TR) on the dimensions of perceived value (PV) on mobile phone enabled internet services (MPEIS). Five dimensions of PV, namely utilitarian value, hedonic value uniqueness value, epistemic value, and economic value were considered in this study. A sample of 522 adult mobile phone users were selected covering the Central Province of Sri Lanka using the multistage random sampling. Confirmatory factor analysis was used to validate the measurement model. Covariance based structural equation modelling was used to analyse the structural model. This study revealed that users’ TR has a positive influence on the epistemic value of MPEIS, and a negative influence on the economic value of MPEIS. It was also evident that users’ TR is not influenced by the utilitarian value, hedonic value and uniqueness value of MPEIS.


INTRODUCTION
The importance of the service industries is growing worldwide and both the industry experts and scholars are interested in developing the knowledge base of service consumers to enable the delivery of services with higher value for the consumer (Yieh et al., 2012). Mobile phones and the mobile services have been revolutionized by the recent developments in the technology. Especially, by the technologies that enabled the transmission of data using a mobile phone (Vriendt et al., 2002). This technological shift was demarcated in the early 1990s with the transformation of mobile communication technologies from analogue to digital. By the year 2003, the mobile phones with accessibility to email services were released to the market. Since then, the mobile technologies have been developed to provide better services in terms of data usage (Lee, 2014). Thus, the mobile phone had transformed into a multi-tasking and multi-purposive device which has provided unlimited growth potentials for the Mobile Service Providers (MSP) (Vriendt et al., 2002).Customers using mobile phones and services have many individual differences. Main objective of this study is to investigate the influence of users' Technology Readiness (TR) on the dimensions of Perceived Value (PV) of mobile phone enabled internet services (MPEIS). The existing literature contains only a limited knowledge about how individuals perceive the technologies in a setting where the use of technology is voluntary and the users are charged for the usage (Wang and Yu, 2014). In this type of a setting PV plays a major role among the many determinants of the usage decisions (Hsu and Lin, 2016).
Despite the multidimensional nature of the PV, a majority of researches have conceptualized it as a single dimension (Al-Debei and Al-Lozi, 2014). There is an inadequacy in the existing knowledgebase to explain the influence of the dimensions of perceived value on the use of technologies in the consumer context (Xu et al., 2015). This study has adopted the five dimensions of PV proposed by Al-Debei and Al-Lozi (2014) in the context of mobile services namely: (1) Utilitarian Value, (2) Hedonic Value, (3) Uniqueness Value, (4) Epistemic Value and (5) Economic Value. Utilitarian Value refers to the extent of effectiveness and efficiency that is perceived by consumers when using information systems (Kim and Han, 2009). Hedonic Value is defined as the level of pleasure and joy users experience when using a certain technology (Al-Debei and Al-Lozi, 2014). Uniqueness Value refers to the sense of differentiation or distinctiveness from others (Tian and McKenzie, 2001). Epistemic Value refers to the knowledge gained upon trying new things (Pihlström and Brush, 2008). Economic Value refers to the gap between value perceived from using mobile internet and the monetary sacrifices (Heinonen and Andersson, 2003). Technology readiness has been defined as "people's propensity to embrace and use new technologies for accomplishing goals in home life and at work". The construct reflects "an overall state of mind resulting from a gestalt of mental enablers and inhibitors that collectively determine a person's predisposition to use new technologies" (Parasuraman, 2000). Parasuraman (2000) has predicted that there can be a relationship between users' TR and PV of MPEIS. This association has not been sufficiently investigated in the existing literature (Yieh et al., 2012). Addressing this knowledge gap five hypotheses were tested in the study. They tested whether the users' technology readiness has a significant influence on the  Conceptual framework of the study is depicted in figure 1 above. The relationship between users' technology readiness with the PV of mobile phone enabled interned services was examined in the study. The MPEIS was analysed in five dimensions. The impact of demographic factors has not been examined in the context of this study.

METHODOLOGY
The population of the study was adult (above 18 years) mobile phone users of the Central Province of Sri Lanka. Central Province was selected due to two reasons: (1) the mobile phone penetration rate of the province was the closest value to the mean penetration value of the country and it represented the median value among the other provinces and, (2) the residents' profile of the province was represented by all the major groups of the population (Central Bank of Sri Lanka, 2015).
Obtaining the sample frame for this study was restrained due to the regulations of the mobile telecommunication industry. Therefore, a multistage random sampling approach based on the administration units of the Central Province was employed to derive a representative sample for the study. Administratively, Central province is divided in to three districts: Kandy, Matale and Nuwara Eliya. Each district is divided in to a number of Divisional Secretariat Divisions and in the lowest level, the province is divided into a total of 2,224 Gramaniladhari Divisions (GND). Two stages of the sampling strategy and their justifications are presented below.

Stage 1
As the initial stage of the sampling process, each district was assigned a quota of GNDs proportionate to the adult population of each district (Table 1). Due to the cost and time constraints, data collection was conducted only in 55 randomly selected GNDs. In stage 2 of the sampling process, 10 respondents were identified from each selected GND using purposive sampling technique. This technique was adopted due to the non-availability of the sampling frame at the GND level as well. The final sample would still be a representative sample of the province since the survey was carried within the representative sample of GNDs selected under Phase1. The sample of 550 mobile phone users was in line with the guidelines given by Krejcie and Morgan (1970). They have indicated that, for populations above the size of 1,000,000, sample should consist of at least 384 units when a confidence level of 95% with a margin of error of 5% is considered. Moreover, this satisfies the requirements of Covariance Based Structural Equation Modelling (Hair et al., 2011). After removing the incomplete responses, 522 responses were available for the analysis, which included 258 responses from males and 264 responses from females. The education profile of the respondents revealed that 371 respondents have at least passed GCE Advanced Level examination, while 22 respondents indicated to have educational qualifications below GCE Ordinary Level. The employment profile of the sample revealed that 109 of the respondents were unemployed, 34 of the respondents were self-employed, and the remaining respondents were employed during the time of data collection.

Measurement of variables
Users' TR has been measured using "The Technology Readiness Index 2.0" (TRI 2.0) which was developed by Parasuraman and Colby (2014). Since this scale is copyrighted by A. Parasuraman and Rockbridge Associates, written permission was obtained from the authors to use the measurement in this study. The scale used four dimensions namely: (1) Optimism, (2) Innovativeness, (3) Discomfort and (4) Insecurity. As per the guidelines from the authors, the items under insecurity and discomfort dimensions were reverse coded since they represent negative themes. Thereafter, TR was calculated by averaging the scores obtained for the four dimensions (i.e. innovativeness, optimism, insecurity and discomfort). The scale developed by Al-Debei and Al-Lozi (2014) was adopted to operationalise the five dimensions of PV. The measurement scale contained three or more items for each dimension and each item was measured by a 5 point Likert scale.
This study investigated the causal relations between a set of variables. The commonly used analytical techniques to investigate causation can be identified as factor analysis, path analysis, structural equation modelling and regression (Sekaran and Bougie, 2013). Structural Equation Modelling (SEM) was employed as the main data analysis technique for the study. The interrelationships among multiple variables were evaluated through regression analyses (Henseler et al., 2016). Literature provide guidelines to select the most appropriate technique between variance and covariance based SEM techniques. In a situation where the objective of the study is to test and confirm an existing theory, covariance based SEM (CB-SEM) would be the appropriate technique. (Hair et al., 2011(Hair et al., , 2013(Hair et al., , 2014. Since this study focuses on theory testing, the CB-SEM has been selected as the appropriate analysis technique. AMOS version 23 has been used for CB-SEM analysis. Questionnaires were distributed to a sample of 30 mobile phone users in the same province and the data were analysed to determine the reliability of the items. Results indicated that, Cronbach's Alpha values for all the latent variables were higher than 0.7 indicating a reliable survey instrument. Questionnaire was translated in to Sinhala and Tamil since the majority of the potential respondents were users of native languages. Traditional forward and backward approach has been followed in the translation process (Degroot et al., 1994) to ensure the same meaning in all three languages.

RESULTS AND DISCUSSION
An Exploratory Factor analysis has been carried out to investigate the item factor loadings. The Principle Component Analysis has been used as the extraction method and Varimax has been used as the rotation method. The items UTV4, HV4, UNV1, and EPV4 have been removed from the analysis to ensure the uni-dimensionality due to the lower factor loadings than 0.6 (Ziegler and Hagemann, 2015). The factor loadings of the retained items for further analysis were above 0.7.

Confirmatory factor analysis (CFA)
Prior to testing the inter-relationship among the variables, the unidimensionality, validity, and reliability for each latent variable has been confirmed using CFA. All the latent variables and their measurement items are used in the analysis. The CFA model is presented in Figure 2 and the Table 2   Results presented in Figure 2 indicated that the measurement model has achieved Unidimensionality since all the factor loadings were higher than 0.5 (Kose and Demirtasli, 2012). The lowest factor loading for the model was 0.68. all the AVEs presented in Table 2 were higher than 0.5 indicating an adequate level of convergent validity (Nunnally and Bernstein, 1994).  Table 3. Discriminant validity is achieved when the square root of AVEs of each construct are larger than the correlation of the specific construct with any of the other constructs in the model. The square root of AVEs presented as the diagonal items in Table 3 are larger than the squared correlation between constructs presented as off diagonal items indicating an adequate level of Discriminant validity. As indicated in Table 2, Cronbach's Alpha value of each dimension was higher than 0.7 and the CR value of each dimension was higher than 0.7. Thus, the measurement model indicates an adequate level of internal consistency and reliability (Nunnally and Bernstein, 1994;Hair et al., 2011Hair et al., , 2013.  Table 4 presented the model fit criteria as per the guidelines of Hu and Bentler (1999). Results indicate that all the values achieved for the CFA model were in the acceptable range. Thus, it can be concluded that the model indicated an adequate level of model fit under all categories.

CONCLUSIONS
Study has investigated the influences of TR on the dimensions of PV. It was found that users' TR has a significant positive influence on the Epistemic Value of MPEIS. In other words, users' TR has a positive influence on the knowledge gained through trying new things. It can be concluded that, higher the users' TR, higher is the knowledge gained from trying new things. It was also found that users' TR has a negative impact on Economic Value of MPEIS.
In other words, users TR has a negative influence on the gap between the value perceived from using mobile internet and the monetary sacrifices. It can be concluded that, higher the TR possessed by a user, lessor is the gap between value perceived from using mobile internet services and the monetary sacrifices. Studies of this nature are rare in the existing literature (Erdoğmuş and Esen, 2011;White et al., 2017). Even the existing few studies have not considered the dimensions of PV used in this study. Thus, evidences cannot be found in the literature to support or contradict these findings. However, the findings needs to be verified with further researches on different technologies and consumer groups. The Mobile Service Providers could use these findings to improved customization of their service packages to achieve higher consumer satisfactions.