When the results of a study are in accordance with what the participants say, we say the study is valid. Data are considered valid if there is no clear difference between the conditions reported by the research subjects and the actual conditions experienced by the person. The Product Moment correlation approach is used for the purpose of validity testing. At a significance level of 5%, we compare the item-specific correlation coefficients with the values in the r table. If the calculated r table value is greater than that value, the validity test concludes that the questionnaire is valid. The survey fails if the calculated r-table value is lower than the r-table value.
| Research Variables | Item | r Count | r Table | Sig | Sig Standard | Information |
|---|---|---|---|---|---|---|
| Digital Payment (X1) | X1.1 | 0,772 | 0,1966 | 0,000 | <0,05 | Valid |
| X1.2 | 0,808 | 0,1966 | 0,000 | Valid | ||
| X1.3 | 0,802 | 0,1966 | 0,000 | Valid | ||
| X1.4 | 0,766 | 0,1966 | 0,000 | Valid | ||
| X1.5 | 0,786 | 0,1966 | 0,000 | Valid | ||
| X1.6 | 0,774 | 0,1966 | 0,000 | Valid | ||
| X1.7 | 0,736 | 0,1966 | 0,000 | Valid | ||
| X1.8 | 0,723 | 0,1966 | 0,000 | Valid | ||
| Government Support (X2) | X1.9 | 0,800 | 0,1966 | 0,000 | Valid | |
| X1.10 | 0,812 | 0,1966 | 0,000 | Valid | ||
| X2.1 | 0,727 | 0,1966 | 0,000 | <0,05 | Valid | |
| X2.2 | 0,774 | 0,1966 | 0,000 | Valid | ||
| X2.3 | 0,703 | 0,1966 | 0,000 | Valid | ||
| X2.4 | 0,846 | 0,1966 | 0,000 | Valid | ||
| X2.5 | 0,780 | 0,1966 | 0,000 | Valid | ||
| X2.6 | 0,784 | 0,1966 | 0,000 | Valid | ||
| X2.7 | 0,713 | 0,1966 | 0,000 | Valid | ||
| X2.8 | 0,649 | 0,1966 | 0,000 | Valid | ||
| Islamic Business Ethics (X3) | X2.9 | 0,724 | 0,1966 | 0,000 | Valid | |
| X2.10 | 0,735 | 0,1966 | 0,000 | Valid | ||
| X3.1 | 0,780 | 0,1966 | 0,000 | <0,05 | Valid | |
| X3.2 | 0,794 | 0,1966 | 0,000 | Valid | ||
| X3.3 | 0,737 | 0,1966 | 0,000 | Valid | ||
| X3.4 | 0,770 | 0,1966 | 0,000 | Valid | ||
| X3.5 | 0,787 | 0,1966 | 0,000 | Valid | ||
| X3.6 | 0,770 | 0,1966 | 0,000 | Valid | ||
| X3.7 | 0,803 | 0,1966 | 0,000 | Valid | ||
| MSME Income (Y) | X3.8 | 0,740 | 0,1966 | 0,000 | Valid | |
| X3.9 | 0,813 | 0,1966 | 0,000 | Valid | ||
| X3.10 | 0,784 | 0,1966 | 0,000 | Valid | ||
| YP1 | 0,783 | 0,1966 | 0,000 | <0,05 | Valid | |
| YP2 | 0,774 | 0,1966 | 0,000 | Valid | ||
| YP3 | 0,722 | 0,1966 | 0,000 | Valid | ||
| YP4 | 0,777 | 0,1966 | 0,000 | Valid | ||
| YP5 | 0,769 | 0,1966 | 0,000 | Valid | ||
| YP6 | 0,766 | 0,1966 | 0,000 | Valid | ||
| YP7 | 0,707 | 0,1966 | 0,000 | Valid |
All statement indicators have significant r values lower than 0.05, as shown in Table 1, and the calculated r value is higher than the table r value. All aspects of these indicators are valid. The variables of this study can be measured appropriately using these indicators.
Reliability is determined by how well the results hold up when we measure the same symptoms multiple times with the same instrument. When the reliability coefficient (r₁₁) is more than 0.6, the research instrument is considered reliable according to Cronbach's Alpha.
| Variables | Cronbach's Alpha | Information |
|---|---|---|
| Digital Payment | 0,928 | Reliable |
| Government Support | 0,910 | Reliable |
| Islamic Business Ethics | 0,927 | Reliable |
| MSME Revenue | 0,874 | Reliable |
Table 2 shows that all variables have Cronbach Alpha values higher than 0.6. All research variables have been determined to be reliable based on the criteria mentioned previously.
Many statistical tests exist, varying from basic to complex methodologies, depending on the features of the data and the objectives or challenges of the research.
Creating a graph showing the distribution of scores is a basic way to check for normality. The simplicity of this approach makes the normality test highly dependent on the observer's accuracy in interpreting the graphical representation of the data. If the volume of data is substantial and its distribution deviates from normality, the possibility of error in conclusions increases. The normality test of residual values and the Kolmogorov-Smirnov were used in this study. To reduce the possibility of error, the normality test was supplemented with the Probability Plot test. A detailed explanation of the evaluation process is provided below.
| One-Sample Kolmogorov-Smirnov Test | |||
|---|---|---|---|
| Unstandardized Residual | |||
| N | 100 | ||
| Normal Parametersa,b | Mean | .0000000 | |
| Std. Deviation | 1.64239639 | ||
| Most Extreme Differences | Absolute | .075 | |
| Positive | .049 | ||
| Negative | -.075 | ||
| Test Statistic | .075 | ||
| Asymp. Sig. (2-tailed)c | .177 | ||
| Monte Carlo Sig. (2-tailed)d | Sig. | .171 | |
| 99% Confidence Interval | Lower Bound | .161 | |
| Upper Bound | .181 | ||
| a. Test distribution is Normal. | |||
| b. Calculated from data. | |||
| c. Lilliefors Significance Correction. | |||
| d. Lilliefors' method based on 10000 Monte Carlo samples with starting seed 2000000. | |||
The following is the result of the One Sample Kolmogorov-Smirnov test. Compared with the 0.05 limit, the two-sided significance level is 0.177. Based on this result, the residual data from this study follows a normal distribution. The results that pass the normality test follow a normal distribution.
The probability plot for the normality test is illustrated in Figure 3 below:
The second illustration shows a histogram of the trend to the right indicating a downward trend. Both sets of figures show a normally distributed distribution for the residual values used in this analysis.
A number of independent variables in multiple regression analysis are represented by the correlation coefficient (r) value. This phenomenon can be understood by conducting a multicollinearity test. Variance Inflation Factor (VIF) and Tolerance Indicator can be used to detect multicollinearity. The model does not show multicollinearity if Tolerance is greater than 0.10. Once Tolerance drops below 0.10, multicollinearity occurs. Multiple correlation is present when the VIF score is greater than 10. It seems that the independent variables are not multicollinear, because their VIF scores are less than 10.
| Variables | VIF | Tolerance | Multikolinearitas |
|---|---|---|---|
| Digital Payment | 1.018 | 0,982 | Not occur |
| Government Suppor | 1.021 | 0,980 | Not occur |
| Islamic Business Ethics | 1.039 | 0,963 | Not occur |
Based on table 4, here are the test results:
VIF requirements, VIF < 10, conclusion: Multicollinearity does not occur.
X1: 1.018 is less than 10, there is no multicollinearity.
X2: 1.021 is less than 10, there is no multicollinearity.
X3: 1.039 is less than 10, there is no multicollinearity.
Tolerance requirements, Tolerance > 0.10 the conclusion is that there is no multicollinearity.
X1: 0.982 is greater than 0.10, there is no multicollinearity.
X2: 0.980 is greater than 0.10, there is no multicollinearity.
X3: 0.963 is greater than 0.10, there is no multicollinearity.
If there is a change in the absolute residual variation after data processing, then the heteroscedasticity test will determine it. If the assumption of no heteroscedasticity is violated, the interpretation of analytical data is disturbed in both small and large sample sizes, which causes reduced accuracy in estimating the regression coefficients. This study tests heteroscedasticity using scatterplot graphs.
Heteroscedasticity in scatter diagrams can be caused by the following things:
a. Some of the other data are below the zero line and some are above it.
b. Both above and below the zero line, the data points do not appear to be concentrated.
c. The data points do not follow a clear pattern, such as an up or down wave shape.
d. The data points should be randomly distributed, with no clear pattern. Figure 4 is a reference.
Above and below the zero line, the data in Figure 4 are scattered. The distribution of the points does not show any waves, and there is no pattern of expansion, contraction, and expansion. As a result, the research model meets the efficiency requirements for understanding the analysis findings.
By using scatter diagrams and the Glejser test, this study tests the possibility of heteroscedasticity. Assuming all independent variables have a Sig. value higher than 0.05, the Glejser test can eliminate heteroscedasticity from the regression model.
| Coefficientsa | ||||||
|---|---|---|---|---|---|---|
| Model | Unstandardized Coefficients | Standardized Coefficients | t | Sig. | ||
| B | Std. Error | Beta | ||||
| 1 | (Constant) | 1.524 | 1.124 | 1.356 | .178 | |
| Digital Payment | -.022 | .016 | -.138 | -1.366 | .175 | |
| Government Support | .002 | .017 | .010 | .094 | .925 | |
| Islamic Business Ethics | .015 | .016 | .100 | .982 | .329 | |
| a. Dependent Variable: Abs_RES | ||||||
The t-values for the following variables Digital Payment, Government Support, and Islamic Business Ethics are 0.982, 0.094, and -1.366, respectively, according to the Glejser test. The significance levels of digital payment (0.175), Government Support (0.925), and Islamic Business Ethics (0.329) are all more than 0.05. Since all independent variables in this study have a significance level greater than 0.05, the regression equation model does not show heteroscedasticity.
Statistical testing and calculation of test values, evaluation of test results, determination of significance levels, and finally, drawing conclusions are various stages that include the implementation of hypothesis testing in this study. The process begins with the formulation of the null hypothesis (H₀) and the alternative hypothesis (Ha).
1) Partial Test (t)
a) Making a hypothesis in a sentence description
b) Making a hypothesis in a sentence description
H0: Digital Payments, Government Support, and Islamic Business Ethics have an insignificant effect on MSMEs' Income.
Ha: Digital payments, Government Support, and Islamic Business Ethics have a significant effect on MSMEs' income.
Determining the level of significance (α)
c) Testing rules
If, t count < t table, then Ha is rejected while H0 is accepted.
If, t count > t table, then Ha is accepted while H0 is rejected.
d) Comparing t-table and t-count
The calculated t-value is compared with the t-table to ensure the assessment of the proposed hypothesis.
e) Making a decision
The final phase of this testing procedure is to ensure acceptance or rejection of the hypothesis based on the data obtained from the statistical calculations performed.
| Coefficientsa | ||||||
| Model | Unstandardized Coefficients | Standardized Coefficients | t | Sig. | ||
| B | Std. Error | Beta | ||||
| 1 | (Constant) | 9.171 | 1.899 | 4.830 | .000 | |
| Digital Payment | .178 | .028 | .458 | 6.460 | .000 | |
| Government Support | .176 | .028 | .446 | 6.285 | .000 | |
| Islamic Business Ethics | .131 | .026 | .353 | 4.934 | .000 | |
| a. Dependent Variable: MSME Income | ||||||
The regression equation is obtained from Table 6 above as follows:
Partial testing is used to test the independent variables against the dependent variable one by one. The table above shows the results of the t-test calculations for each variable X1, X2, and X3, with a significance value < 0.1, to determine whether they have an effect on the dependent variable Y by comparing the t-table with N = 100 samples and α = 0.05. The t-table value obtained is 1.98498, which results in:
The t-test for the Digital Payment variable (X1) shows a t-calculated value of 6.460 with a significance level of 5%. Since the t-calculated value (6.460) is greater than the t-table value (1.98498) and the significance value is less than 0.05, it can be concluded that, partially, the Digital Payment variable has a significant effect on the MSME Income variable (Y). The decision is that Ha is accepted, meaning that the Digital Payment variable has an effect on increasing MSME income in Medan.
The t-test for the Government Support variable (X2) shows a t-calculated value of 6.285 with a significance level of 5%. Since the t-calculated value (6.285) is greater than the t-table value (1.98498) and the significance value is less than 0.05, it can be concluded that, partially, the Government Support variable has a significant effect on the MSME Income variable (Y). The decision is that Ha is accepted, meaning that the Government Support variable has an effect on increasing MSME income in Medan.
The t-test for the Islamic Business Ethics variable (X3) shows a t-calculated value of 4.934 with a significance level of 5%. Since the t-calculated value (4.934) is greater than the t-table value (1.98498) and the significance value is less than 0.05, it can be concluded that, partially, the Islamic Business Ethics variable has a significant effect on the MSME Income variable (Y). The decision is that Ha is accepted, meaning that the Islamic Business Ethics variable has an effect on increasing MSME income in Medan.
2) Simultaneous Test (F)
By taking into account the total influence of all independent variables in the model, this test finds out whether the variable influences Y. Scheme of the testing process:
a. The dependent variable is MSME Income; the independent variables are digital payments, Government Support, and Islamic Business Ethics. At the same time, please describe the alternative hypothesis (H1) and the null hypothesis (H0) related to these variables.
b. Determine the acceptance and rejection zones in the F table by calculating the degrees of freedom (d = k; n - k) and setting α = 5% or 0.05.
c. The F-value from the table is compared with the F-value calculated using k degrees of freedom in the numerator and n minus k degrees of freedom in the denominator, all at a significance level of 5% (α). The following are the components needed to conduct a hypothesis test simultaneously: The null hypothesis (Ho) is rejected, indicating significance, if the F-count exceeds the F-table at α = 5%. The null hypothesis (Ho) is accepted, meaning statistically insignificant, if the F-count is smaller than the F-table at α = 5%. Here are the results.
| ANOVAa | ||||||
|---|---|---|---|---|---|---|
| Model | Sum of Squares | df | Mean Square | F | Sig. | |
| 1 | Regression | 297.311 | 3 | 99.104 | 35.626 | .000b |
| Residual | 267.049 | 96 | 2.782 | |||
| Total | 564.360 | 99 | ||||
| a. Dependent Variable: MSME Income | ||||||
| b. Predictors: (Constant), Islamic Business Ethics, Digital Payment, Government Support | ||||||
Based on table 7, the F-count is 35.626 > from the F-table 2.70 and the F significance value is 0.00 < 0.05, so the hypothesis can be accepted. This shows that together the independent variables consisting of Digital Payment (X1), Government Support (X2), and Islamic Business Ethics (X3) have a significant effect on the dependent variable in this case, MSME Income (Y).
3) Test of Determination Coefficient
One way to assess how well a regression model captures changes in the dependent variable is to look at the coefficient of determination. A low R2 score indicates that the independent factors do not provide sufficient explanation for the dependent variable. Accuracy improves as the R2 value approaches 1. Thus, table 8 is derived.
| Model Summary b | ||||
|---|---|---|---|---|
| Model | R | R Square | Adjusted R Square | Std. Error of the Estimate |
| 1 | .726a | .527 | .512 | 1.66786 |
| a. Predictors: (Constant), Islamic Business Ethics, Digital Payment, Government Support | ||||
| b. Dependent Variable: MSME Income | ||||
Medan City MSMEs consider Islamic business principles, digital payments, and Government Support. Based on the results of the determination coefficient test, the corrected R square value is 0.512. These independent factors contribute 51.2% of the variance in MSME income in Medan City. The rest is influenced by other variables.
The purpose of this study is to examine the relationship between X, MSME income, and Y, the relationship between digital payments, government assistance, and Islamic business principles. Linear regression analysis is used in this study. The results are presented in Table 9 below:
| Coefficientsa | ||||||
|---|---|---|---|---|---|---|
| Model | Unstandardized Coefficients | Standardized Coefficients | t | Sig. | ||
| B | Std. Error | Beta | ||||
| 1 | (Constant) | 9.171 | 1.899 | 4.830 | .000 | |
| Digital Payment | .178 | .028 | .458 | 6.460 | .000 | |
| Government Support | .176 | .028 | .446 | 6.285 | .000 | |
| Islamic Business Ethics | .131 | .026 | .353 | 4.934 | .000 | |
| a. Dependent Variable: Pendapatan UMKM | ||||||
Using tabular data and multiple linear regression, we can obtain the following equation:
Based on the equation, the following can be explained:
1. The constant a of 9.171 indicates that if the variables Digital Payment, Government Support, and Islamic Business Ethics remain constant, then MSME income will be only 9.171%.
2. The regression coefficient for the Digital Payment variable of 0.178 indicates that if the level of Digital Payment increases by 1%, MSME income will increase by 0.178%.
3. The regression coefficient for the Government Support variable of 0.176 indicates that if the level of Government Support for MSMEs increases by 1%, MSME income will increase by 0.176%.
4. The regression coefficient for the Islamic Business Ethics variable of 0.131 indicates that if the level of Islamic Business Ethics increases by 1%, MSME income will increase by 0.131%.
The results of the study indicate that digital payment elements affect MSME income. Although using a significance level of 0.00, the t-table value of 1.98498 and the t-count value of 6.460 are both higher than the threshold of 0.05. Ha is accepted because digital payments have a major impact on MSME income. Digital Payments contribute to increasing MSME income in Medan City.
Digital Payment is an automated service that facilitates digital transaction authorization using credit cards, debit cards, bank transfers, or electronic money, making it easier to shop online with safe and instant procedures. According to[13], one way to make monetary transactions is through a digital payment system, which utilizes electronic devices such as smartphones, tablets, and laptops. In this system, electronic money or e-money is recognized as a legitimate payment method and is increasingly used by the public. The main benefit of a digital payment system is its ability to facilitate and accelerate financial transactions without the need for cash or credit cards. One example is the implementation of QRIS which has been proven effective in facilitating transactions and increasing MSME income and digital payment systems such as QRIS can significantly increase consumer satisfaction [14]. Thus, MSMEs can increase their income by increasing operational efficiency and expanding market access made possible by the digital payment system [15].
This research is in accordance with research studies [16] with the title "Analysis of the Influence of Social Media and Digital Payment Usage on MSME Income in Medan City During the Covid-19 Pandemic", and also research [17] with the title "The Influence of E-Commerce, Social Media, Digital Marketing and Digital Payment on the Income of Fashion MSMEs in the Digital Era (Case Study at Little Bangkok Tanah Abang Market)". It is known that if the use of Digital Payment increases, it will also affect the Income of MSMEs in Medan City.
Government Support has an effect on MSME income, according to the research results. With a significance level of 0.00 lower than 0.05, the calculated t value of 6.285 exceeds the t table value of 1.98498. Therefore, Ha can be accepted if there is a statistical relationship between Government Support and MSME income. MSME income in Medan City is driven by government subsidies.
Government assistance includes all policies, regulations, incentives, and resources provided by the government to encourage the growth and development of certain sectors, including MSMEs [18]. According to research [6], this assistance can be in the form of cash incentives, subsidies, favorable tax policies, and training programs to improve the competitiveness of small businesses. Effective government assistance can improve the competitiveness of MSMEs through various training initiatives and business mentoring. In addition, the government plays an important role in the current era of contemporary economic digitalization by encouraging the creation of an environment that supports the development of MSMEs.
This study is in accordance with research [19] with the title "Analysis of the Influence of Digital Marketing and Government Support on the Success of MSMEs in Medan City". It is known that if the use of Government Support increases, it will also affect the Income of MSMEs in Medan City.
Such findings indicate a strong relationship between Islamic Business Ethics and MSMEs’ profits. If we choose a significance threshold of 0.00 less than 0.05, we find that the estimated t-value of 4.934 is greater than the t-table value of 1.98498. Since MSMEs’ income is influenced by Islamic Business Ethics, ha must be true. Companies in Medan City that adhere to Islamic business principles are more able to obtain financing.
The principles of sharia law, which underlie Islamic Business Ethics, emphasize the need to be honest, impartial, transparent, and socially responsible in all financial transactions [20]. The core of this view is that there must be harmony between the spiritual and social components of the business world and an emphasis on financial gain [21]. In Islamic Business Ethics, business actors are expected to adhere to principles such as integrity, equity, social accountability, and the prohibition of detrimental behavior, including usury and gharar [22]. Business ethics function as principles in carrying out daily activities based on integrity, transparency, and professionalism [23]. When applied to the MSME sector, this ethic has the potential to increase consumer trust and loyalty, which in turn can increase company revenue [24]. Therefore, the application of Islamic business principles to MSMEs in Medan City can be an important aspect in building a sustainable and highly competitive company.
This study is in accordance with research [25] with the title "The Influence of Islamic Business Ethics and Creativity on the Profits of Micro, Small and Medium Enterprises (MSMEs) in Boyolali Regency", and also research [26] with the title "The Influence of Islamic Business Ethics on Profits in Micro and Small Businesses in Watang Sawitto District, Pinrang Regency" It is known that if the implementation of Islamic Business Ethics increases, it will also affect the Income of MSMEs in Medan City.
This study found that in Medan City, MSME income (Y) is significantly influenced by digital payments (X1), government assistance (X2), and sharia business principles (X3). Using a significance threshold of 0.00 and an F value of 35.626, the results of the SPSS simultaneous F test evaluating the hypothesis validated the findings.
If the estimated F value (35.626 > 2.70) is greater than the F table value (0.00 < 0.05) and the significance value is less than 0.05, then the independent factors simultaneously influence the dependent variable.
Digital Payments (X1), Government Support (X2), and Islamic Business Ethics (X3) contributed 51.2% to MSME Income (Y) in Medan City. On the other hand, the remaining 48.8% (100% - 51.2%) was influenced by several variables, including company innovation, market access, marketing strategy, capital investment, competitive landscape, and other external influences.
In addition, there is a favorable correlation between Digital Payments (X1), Government Support (X2), and Islamic Business Ethics (X3) with MSME Income (Y). MSME income in Medan City will increase if one of these independent factors is increased.