Abstract Purpose – The purpose of this study is to empirically examine the factors influencing consumer behavioral intention (BI) to use cryptocurrency as a medium of transaction. Constructs from the unified theory of acceptance and use of technology model and an added variable, perceived risk (PR), are examined to predict BI. Age and gender as moderators are retained in this model. Design/methodology/approach – An online survey was used to gather the respondents’ responses on a five-point Likert scale. G * Power was used to calculate the required minimum sample size. A non-probability sampling technique was used to gather data from the 290 respondents based in Malaysia. The final data set was analyzed using the statistical package for the social sciences and SmartPLS software using structural equation modeling. Findings – The results show that three of the five proposed factors (performance expectancy, effort expectancy and facilitating condition) are significant predictors of BI to adopt cryptocurrency as a medium of transaction. Interestingly, PR is not a significant predictor even though prior research studies showed otherwise. Likewise, the relationship between BI and social influence became significant only when age is added as a moderator. Practical implications – Malaysians are still wary of cryptocurrency, even though global tech firms such as Amazon and Microsoft are already accepting Bitcoin as a payment method. This study aims to provide relevant authorities and businesses (i.e. central bank, retail merchants and cryptocurrency exchangers) insights toward understanding the factors consumers focus on if they were to use cryptocurrency as a medium of transaction. Originality/value – Most cryptocurrency research are done in developed countries (i.e. USA, UK and EU) perspective. This research addresses the lack of quantitative literature on significant factors influencing BI to use cryptocurrency in developing country context while taking a PR, age and gender into consideration.
E P y c n a t c e p x e e c n a m r o f r e P 4 4 9 0 . 4 9 0 1 . 8 9 5 2 . 2 E P 1 3 9 0 . 4 4 1 1 . 7 3 6 2 . 3 E P 6 1 9 0 . 9 1 1 1 . 1 2 0 3 . 4 E P 9 4 9 0 . 7 6 9 0 . 8 0 9 0 . 9 2 9 0 . 6 1 1 1 . 3 6 8 2 . 1 E E ) E E y c n a t c e p x e t r o f f E 6 1 9 0 . 9 4 1 1 . 9 3 7 2 . 2 E E 8 9 8 0 . 5 0 1 1 . 5 2 8 2 . 3 E E 3 6 8 0 . 7 0 9 0 . 8 0 7 0 . 8 3 9 0 . 8 6 9 0 . 3 4 0 1 . 8 2 1 1 . 6 1 3 2 . 3 9 3 2 . 4 E E 1 I S ) I S ( e c n e u n i l a i c o S 2 5 9 0 . 6 7 0 1 . 3 2 4 2 . 2 I S 6 2 8 0 . 0 1 2 1 . 2 6 6 2 . 3 I S 1 1 9 0 . 4 4 9 0 . 9 4 8 0 . 5 5 8 0 . 8 5 1 1 . 6 5 5 2 . 1 C F ) C F ( n o i t i d n o c g n i t a t i l i c a F 3 5 8 0 . 4 1 1 1 . 9 5 8 2 . 2 C F 1 3 8 0 . 3 0 1 1 . 2 0 9 2 . 3 C F 5 3 9 0 . 4 4 1 1 . 2 6 4 3 . 4 C F 6 2 9 0 . 3 5 9 0 . 1 7 8 0 . 2 2 9 0 . 6 6 1 1 . 2 3 6 3 . 1 R P
id: 937c1ea0c5cc2ef53799c9a251958f2a - page: 14
R P ( k s i r d e v i e c r e P 7 0 9 0 . 2 5 1 1 . 0 9 5 3 . 2 R P 0 4 9 0 . 9 8 0 1 . 5 5 3 3 . 3 R P 1 1 1 1 5 9 0 . 6 5 1 1 . 0 1 4 3 . 1 I B ) I B ( n o i t n e t n i l a r o i v a h e B 8 0 9 0 . 3 8 1 1 . 5 2 3 3 . 2 I B 0 0 0 1 . 8 3 0 0 1 . 6 2 0 5 3 . 3 I B Behavioral intention 183 Table 2. PLS-SEM assessment results of reflective measurement models BL 34,2 184 Table 3. Discriminant validity using HeterotraitMonotrait ratio (HTMT) and HTMT Inference testing and age were also shown to impact all three models signicantly. Subsequent results showed Gender (cid:3) PE, Gender (cid:3) EE and Age (cid:3) SI as having positive signicance in predicting BI, while the rest are not signicant. Thus, hypotheses H8H12, H14 and H15 were rejected, and further analysis is needed to verify H6, H7 and H13.
id: f616175ff2e6922d97cca39d0d712974 - page: 14
As the results in Table 4 indicated a signicant impact of gender on PE and EE at a 0.05 level, only these constructs were evaluated for the following analyzes. The interaction effect was tested as per Jeremy Dawsons approach (Kishore and Sequeira, 2016) and the result is shown in Figure 2. PE and EE showed a similar behavioral approach to the intention to use cryptocurrencies: both males and females intention to use will increase if there are higher PE and EE. Further examination shows that the PE effect would be stronger on women, while the EE effect is stronger on men. Thus, the ndings led to the rejection of H6 and H7.
id: 7245e064cc651f019e36a43e07498307 - page: 15
SI was evaluated for the age moderator, as it interacted signicantly with age. Like gender, this construct was evaluated using Jeremy Dawsons approach, and the result is shown in Figure 3. From the graph, it can be concluded that while both younger and older participants do show an increase in the intention to use cryptocurrency when SI increases, the effect is weaker for the older age group (interpreted through the atter slope). Nevertheless, the impact of SI on BI is stronger for older people than younger people. Thus, the ndings support H13. The results of the hypotheses acceptances are concluded in Table 5. Figure 4 only shows the path coefcient for the measurement model and the predictive result without the structural models moderating effect. The dotted line signies the variable with no signicant effect.
id: f8fa3e1d30cabbe916922d6dc2872f35 - page: 15