CHAPTER 4 DATA ANALYSIS
4.2 Reliability Test
Table 9: Result for Central Tendenies
Variables Mean Standard Deviation
Adoption of Cashless Payment 4.366 0.657
Perceived Usefulness 4.256 0.675
Perceived Ease of Use 4.131 0.587
Perceived Security 3.712 0.713
Relative Advantage 4.247 0.668
Social Influence 4.080 0.714
N=199
Source: Developed for the research
The adoption of cashless payment, a dependent variable in this research, showed the highest mean among the variables. It showed 4.366 with a standard deviation of 0.657. Furthermore, the independent variable with the highest mean is perceived usefulness which is 4.256 with a standard deviation of 0.675. On the other hand, the independent variable with the highest standard deviation was the social influence, 0.714, and the mean of 4.247. Next, perceived ease of use has the lowest standard deviation, 0.587 and the mean is 4.131. Relative advantage has the second- highest mean of 4.256 with a standard deviation of 0.675. Lastly, perceived security has a standard deviation of 0.713 and the lowest mean of 3.712.
Cronbach's alpha values for all variables are shown in the table above. Among the variables, cashless payment adoption has the highest value (0.916), which is regarded as great dependability, followed by perceived usefulness of 0.904, which is also considered excellent reliability. Following that, three independent variables with strong interval consistency are present: relative advantage, felt security, and social influence. Finally, perceived ease of use has the lowest Cronbach's alpha score of 0.732. In brief, all variables are regarded as very trustworthy and consistent since their Cronbach's alpha values vary from 0.70 to 0.95.
4.3 Regression Analysis
4.3.1 Spearman’s Analysis Correlation
A nonparametric measure of the degree and direction of the relationship between two variables is the Spearman rank-order correlation coefficient. The variables should be on an ordinal scale (Laerd Statistics, 2013). Since the data distribution is not normal, Spearman's analysis correlation was performed. The correlation coefficient ranges from -1 to +1. A score of 0 indicates no relationship between the dependent and independent variables. A number close to 0 indicates a poor relationship between the variables.
Table 11: Result for Spearman’s Analysis Correlation
Variables Spearman’s Correlation
Adoption of cashless payment vs Perceived
Usefulness 0.724**
Adoption of cashless payment vs Perceived
Ease of Use 0.702**
Adoption of cashless payment vs Perceived
Security 0.449**
Adoption of cashless payment vs Relative
Advantage 0.683**
Adoption of cashless payment vs Social
Influence 0.522**
Source: Developed fro the research
The Spearman Correlation Analysis between dependent and independent variables is shown above. According to the findings, perceived usefulness has the most robust connection between dependent variables. The higher the correlation
coefficient, the more accurate and robust the association between correlated variables.
With 0.819, 0.771, and 0.774, respectively, perceived utility, perceived simplicity of use, and relative benefit show a strong positive link with cashless payment acceptance.
The values for social influence and felt security, on the other hand, are 0.648 and 0.558, respectively. The findings revealed that all independent factors had a substantial positive link with cashless payment acceptance in Malaysia's night market.
It is because of p-value is greater than 0.05.
4.3.2 Multiple Regression
Table 12: Result for Multiple Regression Explanatory
Variables Regression
Standardized Coefficient
Beta
Coefficients
Standard Error Significance
level t
Constant 0.178 0.045 2.021**
Perceived
Usefulness 0.401 0.069 0.000 5.641***
Perceived Ease of
Use 0.226 0.071 0.000 3.570**
Perceived Security -0.040 0.049 0.458 -0.744
Relative advantage 0.296 0.062 0.000 4.691**
Social Influence 0.054 0.054 0.355 0.928
DV: Adoption of cashless payment R:0.863
R square: 0.745
Value (less than) <1.5 Not significant (1.5-2) * Alpha value is 0.10 level (2-5) ** Alpha value is 0.05 level
explains 74.5 percent of the adoption of cashless payment. The significance threshold for this research will be 5%, and the goal will be to establish whether or not there is a statistically significant relationship between the independent and dependent variables.
The decision criterion is that if the p-value is lower than 0.05, reject H0 (There is no significant link between the independent and dependent variables); otherwise, do not reject H0. In other words, there is an essential link between the independent variable and the dependent variable.
The findings demonstrated that perceived utility, perceived ease of use, and relative benefit are less than the significance threshold of 0.05. Therefore, these three independent factors have an important link with the adoption of cashless payments.
On the other hand, the p-value for felt security and social impact is greater than 0.05.
Hence, felt security and social impact are not significant between dependent variables.
In summary, perceived security and social impact have been eliminated out of this investigation and the hypothesis outcome at table below.
Table 13: Outcome of the tested hypothesis
No Hypothesis P-value Result
1 H10: There is no significant relationship between perceived usefulness and the adoption of cashless payment.
0.000 Significant 2 H20: There is no significant relationship
between perceived ease of use and the adoption of cashless payment.
0.000 Significant 3 H30: There is no significant relationship
between perceived security and the adoption of cashless payment.
0.458 Not significant 4 H40: There is no significant relationship
between relative advantage and the adoption of cashless payment.
0.000 Significant 5 H50: There is no significant relationship
between social influence and the adoption of cashless payment.
0.355 Not significant Source: Developed for the research
4.3.3 Multi Regression after removed insignificant
Table 14: Multiple Regression After Removed Insignificant Variables Explanatory
Variables Regression
Standardized Coefficient
Beta
Coefficients Standard
Error
Significance level t
Constant 0.175 0.045 2.084**
Perceived
Usefulness 0.412 0.065 0.000 6.148***
Perceived Ease
of Use 0.230 0.070 0.000 3.656**
Relative
advantage 0.295 0.057 0.000 5.057**
R:0.863
R square: 0.743
Value (less than) <1.5 Not significant (1.5-2) * Alpha value is 0.10 level (2-5) ** Alpha value is 0.05 level
>5 *** Alpha value 0.01
Source: Developed for the research
The table above has excluded perceived security and social influence because these two variables have insignificant relationships with the dependent variable.
Hence, this table can conclude that only perceived usefulness, ease of use, and relative advantage have a significant relationship between dependent variables. The coefficient of correlation (R) is 0.863. It shows that the independent variables are highly correlated with the dependent variable. Then, the R square is 0.743. In other words, 74.3% of the dependent variable are explained by these three independent variables.
Equation 4.1
ACP=β+β PU+β PEOU+β RA
4.4 Residual Analysis
4.4.1 Normality Test
Based on the normality test, the Jarque-Bera (JB) value calculation did not equal zero, which shows that the residuals of this model are not normally distributed.
Furthermore, the p-value is smaller than 0.05. Therefore, reject H0. Residuals are not normally distributed.
4.3.2 Heteroscedasticity Test
The p-value for the Breusch-Pagan test is 0.87. The p-value is bigger than alpha 0.05. Hence, do not reject H0. Residuals are homoscedasticity.
Table 15: Heteroscedasticity Test
Sig df Chi-square
0.878 5 2.42
Source: Developed for the research
4.3.3 Multicollinearity Test
The Variance Inflation Factors (VIF) values are less than ten, which suggests that there is no statistical evidence to support the existence of a linear connection among the independent variables. Hence, do not reject H0. There is no multicollinearity among the variables in the model.
Table 16: Multicollinearity Test
Variables Tolerance VIF
Perceived
Usefulness 0.262 3.823
Perceived Ease of
Use 0.331 3.018
Perceived Security 0.460 2.175
Relative Advantage 0.332 3.010
Social Influence 0.390 2.565
Source: Developed for the research
Table 17: Summary Table for Residual Analysis
Diagnostic Hypothesis Decision
Normality Test H0: Residuals are normally distributed
H1: Residuals are not normally distributed
JB failed to achieve zero value, Reject H0. Residuals are not normally distributed.
Heteroscedasticity Test (Breusch- Pagan)
H0: Residuals are
homoscedasticity
H1: Residuals are
heteroscedasticity
p-value >0.05. Do not reject H0. Residuals are homoscedasticity.
Multicollinearity test (Variance Inflation Factor)
H0: There is no multicollinearity among variables
H1: There is multicollinearity among variables
VIF <10. Do not reject H0.
There is no multicollinearity among variables.
Source: Developed for the research
4.4 Conclusion
The chapter utilized the data to describe descriptive analysis, Cronbach’s alpha reliability test, Spearman’s Correlation Analysis, multiple regression analysis and residual analysis. In the next chapter, the researcher will further describe the main results and conclusion of the investigation.