Research Results On Tax Compliance: Demographics, Reliability Test, Factor Analysis And Regression

Demographics of the Respondents

Discuss about the Case Factors Influencing An Individuals Income Tax.

Results shows that majority of the respondents are female and the proportion of male comparatively less. Usually in the tax payer database male dominate female taxpayers. Similarly the results from age distribution shows that 44 % of the respondents are in the age group of 30 -40, followed by 32 % in 20 -30 age group. In terms of education level 51 5 have bachelor degree followed by 47 % with master degree. Very less respondents have Phd degree and also the respondents with only high school degree is also less. Around 58 % of the respondents are studying accounts which means that people are trying to get the knowledge about the tax compliance.

In terms of position in the most of the respondents are in the middle management followed by high management and executive level. This shows that the respondents in the study are most likely to be in the tax paying bracket. When asked about the TRN number 86 % of the respondents said that they have TRN number which indicates most of the them pay their tax, .Results from the tax registration type shows that most of them comes under company VAT registration followed by legal person VAT registration. Results from turn over shows that majority of the firms have turnover between 375000 and 1.5 million followed by 27 % having turnover between 187000 and 375000. The detailed table and graph of the demographic profile has been shown in the appendix 1.

For the current research the reliability test has been conducted to examine the internal consistency of the questions included in the study. The reliability has been examined using the Cronbach’s alpha. Since the questionnaire included the liker scale questions, it is considered to be the most appropriate measure of reliability. In the current research the liker scale was set as following:

1-Strongly Disagree/ Strongly Unlikely

5- Strongly Agree/Very Likely

Generally the value of Cronbach’s alpha above 0.6 is considered to be acceptable. The results for Cronbach alpha in the current case is shown in the table below and its shows that the value is 0.907. This indicates that the questions included in the study to collect the data as per the requirements of the study.

Reliability Statistics
Cronbach’s Alpha	N of Items
.907	36

Table 1 Results from the Cronbach’s Alpha

While conducting the reliability test, it was also analyzed whether removing some variable can further lead to increase in the Cronbach’s Alpha value.

Reliability Statistics
Cronbach’s Alpha	N of Items
.910	35

Reliability Test Results

Table 2 Results from the improvised Cronbach Alpha

On the basis of the results it was realized that, if the variable (question no 31) is removed from the variable list the value of the Cronbach’s Alpha increases slightly to 0.910. Even though the value do not increase much, it is still better than the value when all the variables are included.

Reliability test for each independent variable

Reliability Statistics
Cronbach’s Alpha	Cronbach’s Alpha Based on Standardized Items	N of Items
.872	.876	6

Summary Item Statistics
	Mean	Minimum	Maximum	Range	Maximum / Minimum	Variance	N of Items
Item Means	4.200	4.013	4.450	.437	1.109	.040	6
Item Variances	.807	.449	1.053	.604	2.345	.045	6

For tax compliance variable the Cron bach alpha is 0.876 which shows that the internal consistency of the factors has been established.

Reliability Statistics
Cronbach’s Alpha	Cronbach’s Alpha Based on Standardized Items	N of Items
.727	.731	6

Summary Item Statistics
	Mean	Minimum	Maximum	Range	Maximum / Minimum	Variance	N of Items
Item Means	4.010	3.854	4.132	.278	1.072	.011	6
Item Variances	1.245	.860	1.859	.998	2.160	.124	6

For perceived tax compliance control the Cronbach alpha value is 0.727 which is higher than the minimum required of 0.6 however the value is lower as compared to the reliability  of the other independent variables.

Reliability Statistics
Cronbach’s Alpha	Cronbach’s Alpha Based on Standardized Items	N of Items
.896	.893	14

Summary Item Statistics
	Mean	Minimum	Maximum	Range	Maximum / Minimum	Variance	N of Items
Item Means	3.876	3.649	4.199	.550	1.151	.029	14
Item Variances	1.126	.778	1.525	.747	1.960	.063	14

For the last independent variable also the Cronbach alpha is 0.896 which suggests that the internal consistency of the independent variable have also been established.

In this section the results from the factor analysis has been discussed. Factor analysis shows whether the questionnaire used to collect the data has been as per the requirement of the research or not.

KMO and Bartlett’s Test
Kaiser-Meyer-Olkin Measure of Sampling Adequacy.	.347
Bartlett’s Test of Sphericity	Approx. Chi-Square	8552.717
df	630
Sig.	.000

Results from the test shows that the KMO value is 0.374 with the chi square value of 8552.717. The value is statistically significant which indicates that further analysis can be conducted.

On the basis of the screen plot it can be said that there are 12 variable which are significant as the line becomes horizontal from the 13^th variable onwards. Once the Eigen values are less than 1 the lines become horizontal. It should be noted that there are total 36 variable which were originally included in the analysis.

Total Variance Explained
Component	Rotation Sums of Squared Loadings
Total	% of Variance	Cumulative %
1	5.500	15.277	15.277
2	5.464	15.176	30.454
3	3.364	9.345	39.799
4	3.298	9.161	48.960
5	2.933	8.148	57.108
6	2.856	7.933	65.040
7	2.421	6.726	71.766
8	2.176	6.044	77.811
9	1.985	5.513	83.324
10	1.497	4.159	87.483
Extraction Method: Principal Component Analysis.

Table 4 Results from the total variance explained.

As shown in the screen plot, only 12 variables out of  36 variables were able to explain most of the variance as the Eigen value of these variables was higher than 1. The table above shows the variance explained by each of the 12 variables. The first factor is able to explain 15.27 % of the total variance followed by 15.17 by the second variable. Results from the variance explained is shown in the table. Furthermore results also shows that 12 factors are able to explain 87.43 % of the total variance. Rest of the 24 variables are only able to explain 12 % of the variance. The detailed results of the factor analysis has been shown in the Appendix II.

Rotated Component Matrixa
	Component
1	2	3	4	5	6
1		.795
2		.864
3		.552			.652
4				.658
5				.704
6				.627
7	.700
8				.674
9		.730
10			.558
11					.928
12					.906
13	.549
14	.560
15				.822
16			.870
17						.854
18						.704
19	.763
20			.885
21						.813
22	.545		.643
23	.858
24	.627
25			.765
26	.521		.531
Extraction Method: Principal Component Analysis.  Rotation Method: Varimax with Kaiser Normalization.
a. Rotation converged in 9 iterations.

Factor Analysis Results

Furthermore the results from the rotted component matrix shows that the major factors are related to F7, F 13, F14, F 19, F 22, F 23, F 24, and F 26. This is because these factors are in the first column of the matrix which suggest these are the most important factors. Similarly the second important factors include F 1, F2, F3 and F 8 and F 9 as these factors are in the second column of matrix. Similarly the 4^th important factors comes out to be F 4, F5, F6 F8 and F 15. Furthermore F 3, F 11, F 12 are the factors which are the 5^th important factors. Lastly the 6^th important variables include the F 17, F 18 and F 21. This results indicate that all the variables can be categorized in 6 different factors which then can be used to find the impact on the dependent variable.

The regression analysis is has been performed to find the impact of the independent variables on the dependent variable. The impact is decided on the basis of the regression coefficient of the independent. If the coefficient is positive then the independent variable have positive and if the regression coefficient is negative then the impact is also negative. For the current research the dependent and the independent variables are as follows:

• Tax compliance intention

• Tax complying attitude
• Normative expectation
• Perceived Tax compliance Control

• Perceived Tax compliance Control * tax compliance intention

However, before performing the regression analysis it is important to test the assumptions of the linear regression analysis. One of the most important assumption is of normality. To examine the normality, the histogram has been drawn for the dependent variable and the figure is shown above. The results shows that the histogram is bell shaped which confirms that the dependent variable follows the normal distribution. The normality suggests that most of the data points are near the mean value and the variance is low. Since the normality has been established further analysis can be conducted.

Results from the regression analysis are shown in the table below.

Model Summaryb
Model	R	R Square	Adjusted R Square	Std. Error of the Estimate	Durbin-Watson
1	.543a	.295	.275	.42603	2.046
a. Predictors: (Constant), Moderator, Normative , Perceived, Tax compling attitude
b. Dependent Variable: Dependent

Table 5 Results from the Model summary

As shown in the table above the R squared value is 0.29 which suggests that the independent variables included in the data set are able to explain 29 % of the variation in the dependent variable.  Rest of the variation in the dependent variable is explained by some other factors which are not the part of the regression equation.

ANOVAa
Model	Sum of Squares	df	Mean Square	F	Sig.
1	Regression	11.077	4	2.769	15.257	.000b
Residual	26.499	146	.182
Total	37.576	150
a. Dependent Variable: Dependent
b. Predictors: (Constant), Moderator, Normative , Perceived, Tax compiling attitude

Table 6 Results from the ANOVA table

Regression Analysis Results

Similarly the ANOVA table indicates that the F value is 15.25. This value shows the cumulative impact of all the explanatory variables used in regression analysis on the dependent variable. In the current case the F value is statistically significant as the p value is less than 0.05. So the cumulative impact is significant.

Coefficientsa
Model	Unstandardized Coefficients	Standardized Coefficients	t	Sig.	Collinearity Statistics
B	Std. Error	Beta	Tolerance	VIF
1	(Constant)	1.717	.312		5.496	.000
Tax compling attitude	.145	.059	.213	2.454	.015	.640	1.563
Normative	.103	.057	.155	1.791	.075	.649	1.541
Perceived	-.049	.052	-.077	-.945	.346	.721	1.387
Moderator	.350	.078	.365	4.464	.000	.720	1.388
a. Dependent Variable: Dependent

Table 7 Results for regression coefficients

Results from regression coefficient shows that tax compiling has positive and significant impact on the dependent variable. Similarly normative and perceived quality al so shows positive impact on the dependent variable, however the coefficients are not statistically significant. Finally the moderator variable shows statistically significant and positive impact on dependent variable. The regression coefficient of the independent variables can be interpreted as follows;

Tax complaining attitudes: with one unit increase in the tax complaining attitudes, the tax complaining attitude increase by 0.145 units, holding all other factors constant.

Normative expectation – with one unit increase in the normative expectation, the tax compliance intention increases by 0.103, holding all other factors constant.

Perceived tax compliance – with one unit increase in the perceived tax compliance control, the tax compliance intention decreased by 0.049 units, holding all other factors constants.

Moderator – with one unit increase in the moderator variable, the tax compliance intention increased by 0.35 holding all other factors constant.

To examine the existence of the multicollinarity, the VIF test has been performed and the results from the collinearity statistics shows that the VIF values are between 1 and 10 which indicates that there is no problem of multicollinarity in the regression equation(Armstrong, 2012; Iaci and Singh, 2012).

Many previous scholars have also studied the factors affecting the tax compliance intention and different scholars have identified various factors. A study by (MACHOGU, 2011)  have shown that the factors such as public governance quality, ethinicity and the attitudes affect the tax compliance. Another study by (Engida and Abera, 2014) identified different factors such as the economic factors ( benefit and cost of tax evasion, rate of tax, audits), institutional factors ( tax authority, role of government) , social factors (positive perception, government policies) and individual factors ( individual finance, awareness, attitude, knowledge). On the other hand (Remali et al., 2015) analyzed the factors affecting the non-compliance of taxpayers. On the basis of the results from the primary data authors concluded that the tax morale and the personal morale are significantly related to the attitude of tax non-compliance. Furthermore the study by (Marandu, Mbekomize and Ifezue, 2015) review and analyze the factors affecting the taxpayers compliance. Secondary data was collected for the time period 1985-2012, Results from the study suggests that the normative, attitudinal and subjective variables are the major factors affecting the tax compliance. Results from this study are similar to the findings of the current study. This shows that the current research is in line with the previous researches and can help to fill the gap in the existing literature.

Assumptions Testing

Furthermore the results from the data analysis also shows that the tax compliance intention is positively and significantly correlated with tax compliance attitude. Also the relationship between the normative expectation and the tax compliance intention are also positively and significantly correlated. This indicates that people with positive normative expectation about the tax compliance are expected to pay tax. This is may be because the people have positive expectation, that the paid tax will be used for the betterment of the society which will improve the overall wellbeing of all the people. Paying tax is also social responsibility so, people with positive attitude are expected to pay tax on time. In addition the relation between normative expectation and tax compliance attitude is also positively related.

Even though the research tried to cover all the relevant point on the basis of the collected data. However this research also have some limitations. The first limitation is that only the quantitative analysis has been conducted. However for most of the research related to human behavior and attitude qualitative research is considered to be most appropriate. Also the sample size in the current study is also small, so the results from the analysis cannot be generalized to the entire population. Finally the cost and time limitation are also associated with the research.

Taxpayers’ Attitude toward Tax compliance is a good predictor for taxpayers’ intention to comply to the tax law

H1.a:  Behavioral Belief of Taxpayers toward registration compliance toward Tax compliance will influence their intentions toward tax compliance positively.

Since the coefficient if positive the hypothesis is accepted.

H1.b:  Behavioral Belief of taxpayers toward Tax filing and payment on time will influence their intentions toward tax compliance positively

Since the coefficient of the moderator variable is positive and significant the hypothesis is accepted.

H2: Taxpayers’ normative and expectation toward tax compliance is a good predictor for taxpayers’ intention to comply to the tax law

H2.a:  Taxpayers’ subjective norms by other taxpayers and consumers toward tax compliance will influence their intentions toward tax compliance positively.

The coefficient of the normative expectation is positive so the hypothesis is accepted.

H2.b:  Taxpayers’ normative belief the taxpayers on their own toward tax compliance will influence their intentions toward tax compliance positively

The coefficient of moderate variable is positive so the hypothesis is accepted.

H3: Perceived Tax compliance Control of knowledge level toward tax compliance is a good predictor for taxpayers’ intention to comply with the tax law

H3.a:  Tax Awareness toward Tax compliance by taxpayers will influence taxpayer’s intentions toward tax compliance positively.

Since the coefficient of perceived is negative so the hypothesis cannot be accepted.

H3.b:  Tax Knowledge toward Tax compliance will influence taxpayer’s intentions toward tax compliance positively

The coefficient of the perceived is negative so the hypothesis cannot be accepted.

H3.c:  Tax understanding toward Tax compliance will influence taxpayer’s intentions toward tax compliance positively

Since the coefficient of moderator variable is positive and significant, the hypothesis is accepted.

H3.d:  Tax awareness toward Tax compliance is a good mediator for taxpayer’s knowledge toward tax compliance

Since the coefficient of the moderator is positive and significant the hypothesis can be accepted.

H3.E:  Tax knowledge toward Tax compliance is a good mediator for taxpayer understands toward tax compliance

The positive and significant coefficient of moderator variable lead to acceptance of the hypothesis.

H4: Tax intention is a good predictor of tax compliance behavior of taxpayers

The coefficient is positive, however not significant. So the hypothesis cannot be accepted

H5: Training is a good moderator between intention and compliance behavior of tax period

Since the moderator variable is positive and significant so the hypothesis can be accepted.

References

Armstrong, J. S. (2012) ‘Illusions in regression analysis’, International Journal of Forecasting, 6, pp. 689–694.

Engida, T. and Abera, G. (2014) ‘Factors influencing Taxpayers’ Compliance with Tax System: An Empirical Study of Mekelle City, Ethiopia’, Journal of Tax Research 12, 12(2), pp. 433–452.

Iaci, R. and Singh, A. K. (2012) ‘Clustering high dimensional sparse casino player tracking datasets.’, UNLV Gaming Research & Review Journa, 16(1), pp. 21–43.

MACHOGU, P. B. (2011) FACTORS INFLUENCING AN INDIVIDUAL’S INCOME TAX COMPLIANCE: A CASE OF SELECTED MSEs IN KASARANI CONSTITUENCY. Nairobi.

Marandu, E. E., Mbekomize, C. J. and Ifezue, A. N. (2015) ‘Determinants of Tax Compliance: A Review of Factors and Conceptualizations’, International Journal of Economics and Finance, 7(9).

Remali, A. mohd et al. (2015) ‘Factors Affecting the Tax Non-Compliance Attitude’, in National Symposium & Exhibition on Business & Accounting. Tenaga: UNITEN.