Determinants for ar-Rahnu usage intentions : An empirical investigation

This study examines the applicability of theory of reasoned action (TRA) in a context of Islamic pawnshop using structural equation modelling (SEM). The present study presents a simplified theory of TRA, hence it is intended to test whether the two constructs in the theory are acceptable or not in a newly context of Islamic pawnshop. The simplified theory is tested using survey data from 250 respondents. Out of these, only 221 questionnaires are found to be usable whilst the rest are omitted owing to the incomplete responses. The results reveal that attitude was significantly related to the intention to use Islamic pawnshop. Subjective norm was also significantly associated with the intention to use Islamic pawnshop. In sum, the present study provides us valuable insights for service providers to future planning of Islamic pawnshop businesses.


INTRODUCTION
In this study, an Islamic pawn is normally noted as ar-Rahnu.Ar-Rahnu is defined as a method of providing short-term financing to a person by pawning her jewellery to banks or pawnshops as a security.It is one of the micro financing facilities available for low and middle class income earners who are seeking financial assistance to meet two purposes.Firstly, for precautionary purposes, mainly when one encounters unexpected situations such as death and accident which required quick cash, and less cumbersome financial assistance, Islamic pawnshop particularly acts to provide the access to overcome this demanding situation.Secondly, it is to meet one's transaction needs.For instance, the availability of Islamic pawnshop is certainly assisting a small trader to meet her working capital need for the continuity of her business.In addition, according to Ismail and Ahmad (1997), Islamic pawnshop allows individuals to fulfil their financing needs without having to resort to other more expensive means such as loan sharks, moneylenders and interest-based loans.
Gadaian Islam Terengganu (MGIT), followed by Kedai ar-Rahnu Kelantan.Unlike banking institutions, these Islamic pawnshops are more convenient for consumers since they have poor credit and would be denied loans from the banking institutions.This implies that the said institutions offer a quick and convenient way to obtain financial resource.More importantly, the services provided are halal.On account of its merit, the Islamic pawnshop services are now available and ready to access branches of EON Bank, Bank Pertanian Malaysia and Bank Islam Malaysia Berhad (Rosly, 2005).Many banking institutions are expected to join the band wagon in the future.These financial institutions offer Islamic pawnshop services based on the following shariah principles: 1. Qardhul hassan: a financial institution will grant a benevolent financing to the applicant who wishes to pawn his valuable item.The financing will be issued under the concept of qardhul hassan, whereby the applicant is only required to pay the amount borrowed.
2. Rahnu: the applicant is required to place a valuable asset as collateral for the financing provided by the financial institution.
3. Wadiah: the financial institution accepts custody of the valuable asset on a wadiah concept, whereby the financial institution promises to keep the valuable asset in a safe place.Under the concept, the financial institution will charge the applicant for the services rendered in keeping the valuable asset, known as upah.
Owing to increased popularity of the Islamic pawnshop in ummah socio-economic development, there exists a growing interest among scholars in examining the Islamic pawnshops from various perspectives.A considerable number of studies in the literature examined Islamic pawnshop, but most of these studies were from qualitative studies (Rugayah, 1986;Ismail and Ahmad, 1997;Ismail, 2004;Sanusi and Johari, 2006).Rugayah (1986) explained that the customers of pawnshop in Malaysia among others are housewives, and private and public officers who are generally low and middle income earners.It is noted by Rugayah (1986) that most of the consumers used pawnshop in order to meet short-term expenses including for emergency.By the same token, Ismail and Ahmad (1997) noted that the majority of the pawnshop users are those of the low income earners.
Quantitative studies relating to Islamic pawnshop are particularly sparse.To fill this gap, drawing upon the TRA, this study postulates that attitude and subjective norm are essential determinants of Islamic pawnshop usage intention.
The aim of this study is therefore to empirically explore the customers of conventional pawnshops decision to use Islamic pawnshop in the future, and thus, provides groundwork for the purpose of introducing a modified TRA model to better reflect the Islamic pawnshop perspective.

THEORETICAL BACKGROUND AND HYPOTHESES DEVELOPMENT
The study used TRA and opting for the theory is due to two main important reasons.Firstly, the theory has received greater popularity in many disciplines due to the flexibilities possessed by the model which makes it easy to apply to different situations (Ramayah et al., 2009(Ramayah et al., , 2010)).To support this, a study by Ozer and Yilmaz (2011) has successfully extended the theory into a context of accountants.On the other angle, Ramayah et al. (2010) have proven that the theory is applicable to examine the green product purchase intention from perspectives of developing countries.Secondly, the study also deemed to choose the theory because it has been tested empirically by various studies and claimed that the two antecedents of the theory, notably attitude and subjective norm, were universally valid to be extended into various contexts of researches.There were common agreements pertaining to the constructs, and asserted that they are particularly beneficial to predicting individual's acceptance of various situations (Tu et al., 2011).Owing to these merits, the authors have opted for the inclusion of the model in the current research.
The theory was introduced by Fishbein and Ajzen (1975) in order to establish a relationship among beliefs, attitudes, intentions and behaviours (Md. Taib et al., 2008).The TRA suggests that a person's behaviour is determined by his or her intention to perform the behaviour and that this intention is, in turn, a function of the person's attitude and subjective norm toward the behaviour (Fishbein and Ajzen, 1975).The TRA model is based on the premise that humans are rational and that the behaviours being explored are under volitional control (Fishbein and Middlestadt, 1997).
Attitude and subjective norm shape a person's intention to perform a behaviour.After all, a person's intention determines the actually desired behaviour (Ryu et al., 2003).Attributable to its remarkable achievement in developing a behavioural predictive model, the TRA has been applied to a wide variety of research fields including management, marketing, and banking (Ramayah et al., 2003(Ramayah et al., , 2004;;Zainuddin et al., 2004;Ramayah and Mohd Suki, 2006;Md. Taib et al., 2008;Lada et al., 2009).Expounded selectively, the study by Lada et al. (2009) examined customers' perception on the intention to choose halal product in the Malaysia context.In order to explain the intention, the study employed TRA to predict the decision to choose.Zainuddin et al. (2004) examined users and non-users' perception of Islamic banking.These studies however have paid little attention to Islamic pawnshop as a subject of their researches.To counter this, the current research is undertaken to reduce this research gap.
For the purpose, the present study is aimed at simplifying the TRA by only incorporating attitude, subjective norm, and behavioural intention.Attitude and subjective norm are considered as two important constructs which could potentially explain Islamic pawnshop usage intention.To comprehend better the importance of the constructs, not only in previous works but also current work, therefore, the followings provide the discussion of the literature from the two essential determinants.

Attitude
By definition, attitude refers to the evaluative effect of positive or negative feeling of individuals in performing a particular behaviour (Fishbein and Ajzen, 1975).Numerous studies have documented the significant relationship between attitude and intention to use (Ramayah et al., 2003(Ramayah et al., , 2004;;Teo and Pok, 2003;Ramayah and Mohd Suki, 2006).Selectively, Ramayah and Mohd Suki (2006) examined MBA students' intention to use mobile personal computer and found that attitude was significantly related to behavioural intention.In more recent study, Md. Taib et al. (2008) examined musharakah mutanaqisah home financing acceptance among 300 postgraduate students from three public universities in Malaysia namely, IIUM, Universiti Teknologi Mara (UiTM) and Universiti Putra Malaysia (UPM).The study reported that attitude had a significant relationship with behavioural intention.Further, attitude was stronger for Muslim consumers than non-Muslim consumers for musharakah mutanaqisah home financing acceptance.
Further, a study by Gopi and Ramayah (2007) found that attitude had a direct positive impact on intention to use an online trading system.Similarly, in this study, it is expected that the intention to use Islamic pawnshop is particularly influenced by one's attitude.Tarkiainen and Sundqvist (2005) on the other hand, asserted that consumers' intentions to buy organic food can be predicted with their attitude.This finding is in line with a study done by Lada et al. (2009).Lada et al. (2009) examined what factors led to a consumer choice of halal products and found out that "attitude" was significantly related to halal products consumption.The hypothesis below was developed and tested in this study: H 1 : Attitude has a positive effect on the ar-Rahnu usage intentions.

Subjective norm
Subjective norm refers to the individual's perception of the likelihood of the potential referent group or individuals approve or disapprove of performing the given behaviour (Fishbein and Ajzen, 1975).This construct has been applied in different contexts ranging from technologyadoption to non-technology adoption researches.The importance of this construct has been addressed by many scholars such as Ramayah et al. (2004) and Gopi and Ramayah (2007).Explained in more detail, Gopi and Ramayah (2007) found that subjective norm had a direct positive relationship towards behavioural intention to use internet stock trading.Interestingly, a study by Md. Taib et al. (2008) also asserted that subjective norm was an influential determinant in explaining the behavioural intention of postgraduate students on Islamic home financing.Similarly, Lada et al. (2009) found that subjective norm was significantly related to halal products use among consumers.
A study by Tarkiainen and Sundqvist (2005) on Finnish consumers in an organic food buying behaviour, tested the applicability of the theory of planned behaviour (TPB).Tarkiainen and Sundqvist (2005) asserted that consumers' intentions to buy organic food could be predicted with their subjective norm.
Further, some of the previous studies have reported mixed results deviated from the earlier discussed studies.Among others, Chau and Hu (2001) and Lewis et al. (2003) found that subjective norm had no a direct impact on intention to use.On the contrary, Venkatesh and Davis (2000) and Teo and Pok (2003) had reported a significant relationship between subjective norm and behavioural intention.Given these studies, the following Amin and Chong 8183 hypothesis was developed and tested in this study: H 2 : Subjective norm has a positive effect on the ar-Rahnu usage intentions.

Sample
The sample for this study comprised customers of existing conventional pawnshops in Kota Kinabalu, Eastern Malaysia.The participants particularly had experiences in conventional pawnshops.The participants were also local people who had engaged in conventional pawnshops transactions for many years.Worth to address that, the majority of the participants were Muslims as they were mainly Bajau, Malays Suluk, Kadazan-Dusun, and Bugis.Convenience sampling was employed.Considering the sampling method was due to two main reasons.Firstly, the present study, generally, was self-financed in which it was of particular interest to conduct the sampling method in order to overcome this financial constraint.Secondly, the sampling method was considered an easy way to be implemented and it generated time-saving for the present study to be conducted.

Data collection
This study is based on primary data collected through a survey with the help of a questionnaire.The questionnaire was administered by meeting the participants on a one-to-one basis primarily in conventional pawnshops.The questionnaires were hand delivered to the customers to elicit voluntary participation.The survey was conducted between August and October 2009 during working hours.
Respondents were sought individually to complete the survey questionnaire by research assistants who have been specially trained to speed up the process of distributing and collecting of questionnaires.Essentially, the respondents selected, varied according to their age group as well as their willingness to be involved in the survey work.Participants were informed that the participation in the study was voluntary and their responses would be kept confidential and would be analysed only at the aggregate level.From the fieldwork, a total of 250 questionnaires were obtained.A total of 221 usable responses were generated from the 250 questionnaires.

Research instrument
The questionnaire contained two parts.The first part consisted of demographic sample attributes such as a participant's age, education background, level of monthly income and gender.The second part of the questionnaire was devoted to the key issues of the study questions, such as attitude, subjective norm and usage intention.Attitude and subjective norm (the independent variables) were measured using five items each.Similarly, usage intention (the dependent variable) was also measured using five items.A total of 15 items were assessed in the second part.The items of these constructs were adapted from previous researches in various situations: attitude (Lada et al., 2009;Md. Taib et al., 2008); social influence (Md. Taib et al., 2008;Ramayah and Mohd Suki, 2006); usage intention (Gopi and Ramayah, 2007).The items of these constructs were phrased in the form of statements in eliciting a participant's views on the present research subject matter.In all these measures, a Likert-type scale of 1-5 was employed.A score of 1 indicates "strongly disagree" with the statement whereas a score of 5 indicates "strongly agree" with the  Comments suggestions about the question sequence, wording choices, and measures were also solicited, leading to several minor modifications to the questionnaire.These feedbacks were employed to improve the readability and the quality of the questions in the instrument.

Exploratory factor analysis (EFA)
Factor analysis was employed to validate both independent and dependent variable constructs.Principal component analysis with varimax rotation was utilized in all cases.The purpose of using factor analysis is to summarize patterns of correlations among observed variables, to reduce a large number of observed variables to a smaller numbers of factors, and to provide an operational definition (a regression equation) for an underlying process by using observed variables, or to test a theory about the nature of underlying processes (Tabachnick and Fidell, 2007).
Table 1 reports the results of the factor analysis of the independent variables excluding the demographic variables.Table 2 reports the results of the factor analysis of the dependent variable.A factor loading value of 0.60 is regarded as good and significant (Hair et al., 1998).On the independent variables, all 10 items had value of exceeding than 0.60 with 0.853 was the higher.On the dependent variable, all 5 items had value of more than 0.60, which was consistent to that of Hair et al. (1998).It is worth noting that all constructs that achieve eigenvalues greater than one are considered as significant.In the current results, the eigenvalues for the independent  variables were greater than one, as was the dependent variable.It is likely argued that the used items in the constructs were significant and qualified for the further analysis of the study data.
After completing EFA, internal consistency reliability to test unidimensionality was evaluated by Cronbach's alpha.The resulting alpha values ranged from 0.823 to 0.893, which were above the acceptable threshold (0.60) suggested by Hair et al. (1998).In sum, all constructs have good internal consistency reliability.
The data from the EFA were further confirmed using confirmatory factor analysis (CFA).In sum, this EFA gives us confidence to conduct CFA.CFA can be used to assess unidimensionality.A CFA was conducted for each of the 3 constructs to determine whether the 15 items measured the construct they were assigned adequately (Haque et al., 2009).Additionally, satisfactory Cronbach's alpha value did not indicate unidimensionality of a particular scale (Gerbing and Anderson, 1988).Hence, CFA was employed for the assessment of unidimensionality in the present study.

Confirmatory factor analysis (CFA)
Although EFA produced good results, yet, under the CFA, modifications are made to ensure that the hypothesized model meeting the SEM assumptions.Evidently, the items may subject to the modifications, although they were statistically significant under EFA.Considering these modifications are of importance in ensuring the data are free from the outliers and the nonnormality of data.The authors used CFA with AMOS 16 to examine the convergent validity of each construct.The CFA model for each construct is presented in Table 3, 6 and 9.The factor loadings are greater than the recommended level of 0.35, which is based on 221 samples and at the 5% significance level (Hair et al., 1998).
Table 3 presents the CFA for attitude, in which prior to the modification, it was reported that RMSEA was 0.093, which suggests poor fit, and further modification was required.For the purpose, standardized factor loading from the first level of analysis were reconciled to improve the model fit.The constructs of 0.60 were well within the recommended range of acceptability (Hair et al., 1998).The results are presented in Table 4. Attributable to the non-normality, although ATT4 and ATT5 had acceptable values, these items were dropped in order to improve the data normality.Final data are as reported in Table 5. Statistically, since the multivariate normality value was lower than 1.96 (5 percent significance level), therefore, it was concluded that data for attitude were normal and further SEM analysis was valuable.
Table 6 presents the CFA for subjective norm, in which  prior to the modification, it was reported that RMSEA was 0.166, which suggests poor fit, and further modification was required.For the purpose, standardized factor loadings from the first level of analysis were employed to improve the model fit.Following on the recommendation by Hair et al. (1998), all constructs exceeding the cut off value 0.35 were considered significant.Since the items under subjective norm were higher than 0.35, therefore, all of them were significant and retained for further analysis of the data.Unfortunately, retaining all items was not an option owing to the non-normality of data.To counter this, although these items were significant, but due to the non-normality of SN4 and SN5, the option to drop them was of utmost importance in elevating the fittest of data into the model developed.
The decision to drop SN4 and SN5, although significant, was able to improve the multivariate normality, in which, after the deletion of the said items, the data were then found normal as proven by the CR value less than 1.96 (5% significance level).
Table 9 presents the CFA for usage intention, in which prior to the modification, the RMSEA value was 0.203, which suggests poor fit and further modification was required.For the purpose, standardized factor loading from the first level of analysis were analysed to improve the model fit.The items employed under usage intention were found to be reliable as far as the Hair et al.'s (1998) recommendation was concerned.All values ranged from 0.658 to 0.783.Yet, the two items namely USINT4 and USINT5 were dropped in order to improve the normality of data.This approach was done throughout the constructs employed in the present study.
Omitting the two items, as earlier noted, was of significance to improve the RMSEA value and other indices as well.This approach was done to ensure only significant and meaningful items were within in the model employed.It was also importantly done in order to normalise the data to meet the SEM assumption as addressed at the earliest.After the deletion of the two items, the RMSEA value was improved while the data were normally distributed.Table 11 provides the detail.In the table, it is reported that the multivariate value was 0.247 which was lower than 1.96 (5% significance level), therefore, it was concluded that data for usage intention were in the range of normal distribution of data and further SEM analysis was valid to be carried out.Subsequent to the individual modification for a given variable, next stage is to include all the variables employed in order to measure the model fit.Maximum likelihood estimation (MLE) procedure using AMOS is employed.It is a technique, which assumes that the observed variables are continuous and normally distributed (Flora and Curran, 2004).This is particularly efficient since the maximum likelihood method features a number of desirable characteristics such as consistency and asymptotic efficiency (Bauer et al., 2005).The model fit is assessed using RMSEA, RMR, GFI, TLI, CFI and NFI.The cut off value for the goodness of fit indices is based on MacCallum et al.'s (1996) and Byrne's (1998) recommendations.Viewed importantly, the final measurement model fit was considered good after the outliers and nonnormal data were excluded.In other words the final CFA was better.Both RMSEA and RMR were well within the range of acceptability recommended by MacCallum et al. (1996) and Byrne (1998). As MacCallum et al. (1996) recommended, a RMSEA that is less than 0.08 indicates good fit and reasonable errors of approximation in the population.The RMR of the measurement model indicates that the model is a well-fitting model (RMR < 0.05).GFI, TLI, CFI and NFI also suggest that the hypothesised model represented an adequate fit to the data.Worth to mention, the model fit was improved using a conservative strategy, which was none of the error terms is allowed to covary in the measurement model.Further, the freeing of cross-loadings was also not allowed since the existence of significant cross-loading indicated lack of construct validity.

Structural equation modelling (SEM)
SEM is a powerful and flexible analytic method that plays a critically important role in many empirical applications in social science research.Attributable to the general linear model is embedded within SEM, this modelling framework can be used in a wide variety of applications.The general goal of SEM is to test the hypothesis that the observed covariance matrix for a set of measured variables is equal to the covariance matrix implied by a hypothesized model (Flora and Curran, 2004).The nine items as noted in the CFA are further analysed using SEM.As reported in Table 13, both RMSEA and RMR are well within the range of acceptability recommended by MacCallum et al. (1996) and Byrne (1998) respectively.As expounded earlier, MacCallum et al. (1996) suggested that a RMSEA that is less than .08indicates good fit and reasonable errors of approximation in the population.The RMR of the measurement model indicates that the model is a well-fitting model.
Modification indices provide indications of misfit of the structural model.Consequently, errors for the constructs tested are allowed to covary to strengthen the model further.The final SEM for the research model was reported in Table 13.In terms of hypothesis testing, it was found that attitude was significantly related to the intention to use Islamic pawnshop at the 5% significance level (Table 16).Since the two variables have positive relationship, therefore the greater the attitude, the greater the intention to use Islamic pawnshop.The result is in consonance with the study's finding by Lada et al. (2009) who found that attitude had related to the behavioural intention.Subjective norm was also significantly associated with usage intention at the 5% significance level.This means the greater the subjective norm is, the greater the usage intention among participants.This result is in line with the assertions made by Venkatesh and Davis (2000) and Teo and Pok (2003) had reported a significant relationship between subjective norm and behavioural intention.
Particularly, the SEM for the aforementioned demonstrates the distribution of data was normal.Given this evidence, a sample can be considered to be multivariate normally distributed at the 5% significance level since the critical ratio is smaller than that of 1.96, indicating that the coefficient of multivariate kurtosis is not significantly different from zero.This means, the assumption of data normality is particularly met.Hence, the data employed in the present study is statistically meaningful and also helps us to improve the understanding of the subject of the study.
The effect of the two constructs on usage intention is summarized in Table 16.As shown in Table 14, it is reported that attitude was an influential predictor of usage intention and had direct effect on the intention.Similarly, subjective norm was also significant and had direct effect on the intention.By way of comparison, attitude direct effect was greater than that of the subjective norm effect, which means that attitude was relatively the strongest predictor of usage intention instead of subjective norm.

DISCUSSION
The study is aimed at explaining the applicability the TRA in a context of Islamic pawnshop using survey questionnaire to validate it.The model is simplified by only investigating the impact of the "attitude" and "subjective norm" constructs on the usage intentions.The appropriateness of the model's constructs is examined in length using SEM.Since the research is model testing in nature, therefore, there is no additional constructs have suggested in the model.Further, the discussion of the study's finding is presented as follows:

Attitude
In the present study, the attitude was positively correlated to use.Thus, the more positive the attitude is, the more likely that an Islamic pawnshop is selected by the respondents.The result is not surprisingly as many studies have been found out that attitude was an influential factor for behavioural intentions.This outcome is consistent with the findings from (Ramayah et al., 2003;2004;Teo and Pok, 2003;Ramayah and Mohd Suki, 2006).It is confirmed that attitude has strong influence over the behavioral intentions on an Islamic pawnshop.Overall, the results show the appropriateness of the attitude of the TRA in an Islamic pawnshop context.

Subjective norm
The current study's result also concluded that subjective norm was positively associated with use.Importantly, friends, family members and teachers could be the agents to impacting individual's reaction pertaining to Islamic pawnshop use.The findings are in light to that of the previous works opined by Ramayah et al. (2004) and Gopi and Ramayah (2007).These authors have documented that the TRA's subjective norm construct has positive relationship to that of behavioural intentions.The outcome of the current study for subjective norm is also in the direction.This means that, the more positive the subjective norm, the more likely that an Islamic pawnshop is chosen by the respondents of the study.

MANAGERIAL IMPLICATIONS
Managerial implications can be drawn from the present study.Practically, this research assists a service provider in understanding the attitude of existing users of conventional pawnshop in which in the future the service provider is of value to set up with "Islamic pawnshop counters" in their premises in order to cater Muslim financial needs who gives greater emphasis on the religion matters in financial transactions.The research is also pivotal to assist the existing bank-based Islamic pawnshops to pay more attention about the importance of attitude and subjective norm in affecting one's usage intention towards Islamic pawnshop.For the purpose, demonstrating the positive views on the Islamic pawnshop facilities is generally creditable via electronic exhibition such as television.Improved brochures with regard to this issue are also fundamental to promote the use of Islamic pawnshop.On the prospective researcher point of view, the research is of significance to contribute to the knowledge for Islamic pawnshop by providing the latest findings, particularly what makes people using ar-Rahnu.

Conclusion
It is examined whether the TRA is supportive in understanding the impact of attitude and subjective norm on the usage intention, with a reference to the newly emerging context notably Islamic pawnshop.In our opinion, this study is a pioneering effort in applying the TRA to the newly emerging context of Islamic pawnshop.The TRA is evidently very relevant to be applied to the context.This means that the constructs of the TRA are of paramount importance to be adequately generalised into the Islamic pawnshop context.
Every research comes with contributions.As such, the present study is also contributing primarily to the body of knowledge pertaining to Islamic pawnshop.Firstly, it is successfully applied the TRA in a newly context of Islamic pawnshop.The TRA has been sufficiently addressed the impact of the "attitude" and "subjective norm" factors on the Islamic pawnshop use.Secondly, the research is of importance to note that SEM approach is widely scanty for Islamic pawnshop.Various studies have employed SEM in their theoretical frameworks, but particularly on the other occasions (Lada et al., 2009;Tarkiainen and Sundqvist, 2005;Chau and Hu, 2001).The TRA has been tested empirically and received prosperous attentions by Ramayah et al. (2010;2009).For the purpose, the current study is also extended to examine the applicability of the TRA and escalated of our understanding pertaining to the factors facilitating the Islamic pawnshop usage intentions.
In a similar vein, the study also comes with limitations.This study acknowledges two main limitations.The reasons of only suggesting two limitations are of profound importance of developing more works in this area meticulously in the future.Firstly, the present study was conducted in Kota Kinabalu, in which the findings can only be generalised to local people of Sabah.The findings and implications drawn from this present study cannot be readily generalized to other cities in Malaysia.To overcome the limitation, future works may call to include all Malaysians from various states in Malaysia.Including the sample from Kelantan, Terengganu, Johore, Kuala Lumpur, Penang and Sarawak is of value to consider.Secondly, the present study used convenience sampling with the understanding that the current research was self-financed and also owing to the limitation of times.This is only a minor dearth and the authors believe the use of convenience sampling method is able to contribute significantly, at least at exploratory level.In future researches, the selection of different sampling method will be of importance to advance the usefulness of the findings obtained and for the purpose of generalisation of findings.

Table 1 .
Factor loading for independent variables.

Table 2 .
Factor loading for dependent variable.

Table 3 .
Confirmatory factor analysis for attitude.
*Dropped due to non-normality of data.

Table 5 .
Normality of data.

Table 6 .
Confirmatory factor analysis for subjective norm.
*Dropped due to non-normality of data.

Table 8 .
Normality of data.

Table 9 .
Confirmatory factor analysis for usage intention.
*Dropped due to non-normality of data.

Table 11 .
Normality of data.

Table 15 .
Normality of data.

Table 16 .
Path coefficients and strengths of individual paths.