• 沒有找到結果。

CHAPTER 4 RESEARCH RESULTS

4.2 REDEFINING DEMOGRAPHIC FACTORS

4.2.1 Gender information

The statistics describe gender Table 3

Frequency Percent Valid Percent

Cumulative Percent Gender Male

96 47 47 47

Female 109 53 53 100

Total 205 100.0 100.0

In gender group, males account for 47% (96 employees) and females accounted for 53% (109 employees).

34

Figure 5 Gender 4.2.2 Education

Education Statistics Table 4

Frequency Percent

Valid Percent

Cumulative Percent Eduction Low

35 17 17 17

High 170 83 83 100.0

Total 205 100.0 100.0

35

Education was also divided into two groups: group had lower levels (immediate education and colleges) accounted for 17% (35 employees) and those with high levels (university and master) accounted for 83% (170 employees).

Figure 6 Education

4.2.3 Working position

Working position Statistics Table 5

Frequency Percent Valid Percent

Cumulative Percent Working

postition

Management

43 21 21 21

36

Employees 162 79 79 100.0

Total 205 100.0 100.0

The management group accounted 21% proportion with 43 employees and non-management group accounted 79% proportion with 162 employees.

Figure 7 Working position 4.2.4 Working department

Working department Statistics Table 6

Frequency Percent

Valid Percent

Cumulative Percent Department Direct

92 45 45 45.0

37

Indirect 113 55 55 100.0

Total 205 100.0 100.0

The group with employees working in direct department occupied 45%, 92 employees, and the group with employees working in indirect departments occupied 55%, 113 employees.

Figure 8 Department 4.3 SCALE ASSESSMENT

As presented in the chapter 2, the scale factors affecting employee satisfaction in District 2 Tax Department consist of six components: (1) The essence of working, measured by nine variables denoted by V1 through V9; (2) Salary, measured by 8

38

observed variables, denoted V10 to V17, (3) Colleagues, measured by 7 observed variables, denoted V18 to V24, (4) Leadership, measured with 12 observed variables, namely V25 to V36; (5) Training promotion and opportunities, measured by 8 observed variables, represented by V37 to V44 respectively; (6) Working environment, measured by 13 observed variables, denoted V45 to V57 (7) Employee satisfaction in the work of District 2 Taxation department, measured by 6 variables, denoted SET_1 to SET_6. The scale was assessed through the main tool, Cronbach's alpha. Cronbach's alpha is used to remove the garbage variables whose correlation coefficient less than 0.3 will be disqualified.

Criteria for selecting scale is when its Cronbach alpha has reliability that is greater than 0.6 (Nunnally & Burnstein, 1994).

4.3.1 Cronbach Alpha Reliability coefficient

4.3.1.1 Measurement scale of work essence

Results analyzed through SPSS 19.0 for assessing the measurement scale of work essence are shown in Table 7. Values reported reliability coefficient of the component of work essence is 0.827. These corrected item-total correlations are achieved values greater than 0.3. Reported values is reach a low of 0.409 (variable V5) and the highest reported value reached a peak of 0.699 (variable V7).

39

Analyzing the reliability of work essence scale

Cronbach's Alpha N of Items

.827 9

4.3.1.2 Measurement scale of salary

Through analysis we found Cronbach's alpha values reported these corrected item-total correlations have values greater than 0.3. Reported values is lowest as 0.421 (variable V13) and the largest reported value is 0.703 (variable V11).

Cronbach alpha coefficient values reported by 0.834 (detailed in Table 8).

Analyzing the reliability of salary scale Table 8

Scale Mean if

Item Deleted

Scale Variance if Item

Deleted

Corrected Item-Total Correlation

Cronbach's Alpha if Item

Deleted

41

V10

Cronbach's Alpha N of Items

.834 8

4.3.1.3 Measurement scale of colleagues

Evaluating the scale of Colleagues factors is expressed through Table 9. Values reported reliability coefficient of colleagues is 0.936. These corrected item-total correlations are achieved the value greater than 0.3. Reported values is lowest at 0.512 (variable V24) and the highest reported value is 0.762 (variable V19).

42

Analyzing the reliability of colleagues scale

4.3.1.4 Measurement scale of leadership

Cronbach alpha coefficient values reported value of 0.923. These corrected item-total correlations have values greater than 0.3. Reported values is lowest at 0.556 (variable V29) and the largest reported value is 0.714 (variable V36). Details are shown in Table 10.

43

Analyzing the reliability of leadership scale

4.3.1.5 Measurement scale of training and promotion opportunities Analyzing the reliability of training and promotion opportunities scale Table 11

These corrected item-total correlations are achieved values greater than 0.3.

Reported value reached a low of 0.700 (variable V39) and the highest value reported is 0.805 (variable V38). Component training opportunities and advancement has Cronbach's Alpha whose value is .911.

45

4.3.1.6 Measurement scale of working measurement

Exploring table 4.12 saw Cronbach's alpha value was worth 0.927. These corrected item-total correlations have values greater than 0.3. Reported values is low as 0.499 (variable V52) and the largest reported value 0.723 (variable V49).

Details show in Table 12.

Analyzing the reliability of working measurement scale Table 12

V56 V57

43.12 43.45

135.827 134.512

.694 .653

.931 .925

Cronbach's Alpha N of Items

.927 13

4.3.1.7 Measurement scale of Satisfaction

The measurement scale of employee satisfaction in the Tax Department was measured by the six variables. Analyzing via Cronbach Alpha (detailed in Table 4:12), we have achieved Cronbach's alpha value was 0.752. The value of corrected item-total correlations have values greater than 0.3. Reported values is low as 0.459 (variable SET_4) and the largest reported value is 0.66 (variable SET_1).

47

Analyzing the reliability of satisfaction scale

Through the results of Cronbach alpha analysis we found that 6 components of the measurement scale of quality of training are greater than 0.6. Thus, the scale defined in the thesis of statistical is significance and achieve the necessary reliability coefficients. Specifically: (i) Component of work essence whose Cronbach alpha value reached 0.827; (ii) Component of salary had Cronbach alpha value reaching 0.834; (iii) Component of colleagues whose Cronbach alpha value reached 0.936; (iv) Component of leadership occupied Cronbach alpha

48

value reaching 0.923; (v) Component of training and promotion opportunities with Cronbach alpha value reaching 0.911; (vi) Component of working environment possessed Cronbach alpha value reaching 0.927. The measurement of satisfaction had Cronbach alpha value reaching 0.752. So, these six components assessing job satisfaction are used for EFA analysis.

4.3.2 Exploratory Factor Analysis EFA

As discussed in Chapter 3, EFA, Exploratory Factor Analysis, is used to test the measurement scale in the study. The measurement scale in this study include 57 observed variables and after checking the reliability of Cronbach Alpha method there are no variables were excluded. To confirm the suitability of the scale, with 57 observed variables, this study use EFA method. KMO (Kaiser - Meyer - Olkin Measure of Adequacy Simping) index was used to analyze the relevance of these factors. The value of KMO index that greater than 0.5 will be used.

4.3.2.1 Exploratory factor analysis EFA with 6 quality components

H0 hypothesis posed in this analysis is that among 57 variables in population are no correlation with each other. KMO and Bartlett's test in the factor analysis showed that the hypothesis H0 was rejected (sig. = 0.000). KMO coefficient value 0.931 reported greater than 0.5. The EFA results obtained 6 components at eigenvalues that is 1793. The measurement scale is acceptable. However, in 57 observed variables, there are 2 variables (V28, V48) with unsatisfactory values (<0.5). Two variables in the analysis of Cronbach Alpha satisfactory but the analysis did not achieve EFA. Therefore, these two variables will be excluded.

Rotated Component Matrix 1st Table 14

49

Rotated Component Matrix(a)

Component

SET_1 SET_2 SET_3 SET_4 SET_5 SET_6

V3 .763

V4 .708

V5 .704

V6 .630

V7 .618

V1 .585

V2 .568

V8 .560

V9 .548

V12 .753

V13 .702

50

V14 .701

V15 .690

V16 .688

V17 .660

V10 .648

V11 .625

V20 .658

V21 .654

V22 .648

V18 .645

V19 .638

V23 .634

V24 .630

V29 .693

51

V30 .682

V31 .671

V32 .660

V25 .623

V26 .598

V33 .573

V34 .568

V27 .558

V35 .548

V36 .537

V28 .486

V42 .704

V40 .700

V41 .658

52

V43 .630

V44 .618

V37 .598

V38 .558

V39 .537

V49 .754

V52 .738

V53 .729

V45 .663

V46 .652

V47 .650

V57 .646

V50 .620

V51 .614

53

V55 .611

V54 .593

V56 .587

V48 .452

Extraction Method: Principal Component Analysis Rotation Method: Varimax with Kaiser Normalization

The KMO and Bartlett's 1st Table 15

Kaiser-Meyer-Olkin Measure of Samping

Adequacy .931

Bartlett's Test of Approx. Chi-Square 7501.035

Sphericity df 1793

sig. .000

The next step is to proceed to remove the two variables (V28, V48) and then conduct exploratory factor analysis with 55 observed variables: KMO and Bartlett's test the value reported by 0.941 KMO (greater than 0.5), and the meaning value Sig. = 0.000 (<0.05). EFA results obtained from 6 components with the eigenvalues whose value is 1801. From here, the study concluded

54

measurement scale is accepted, the observed variables are correlated with each other considering the overall scope of the sample.

Rotated Component Matrix 2nd Table 16

Rotated Component Matrix(b)

Component

SET_1 SET_2 SET_3 SET_4 SET_5 SET_6

V3 .763

V4 .708

V5 .704

V6 .630

V7 .618

V1 .585

V2 .568

V8 .560

55

V9 .548

V14 .765

V16 .722

V17 .701

V13 .688

V12 .682

V15 .660

V10 .648

V11 .625

V18 .745

V23 .724

V20 .688

V21 .654

V22 .648

56

V19 .638

V24 .630

V27 .724

V30 .713

V25 .690

V34 .682

V26 .623

V33 .598

V36 .587

V35 .568

V29 .563

V31 .558

V32 .548

V42 .714

57

V40 .598

V41 .566

V43 .658

V44 .630

V37 .721

V38 .618

V39 .537

V45 .764

V54 .747

V51 .729

V52 .711

V55 .693

V56 .652

V57 .650

58

V50 .646

V46 .620

V49 .614

V53 .593

V47 .573

Extraction Method: Principal Component Analysis Rotation Method: Varimax with Kaiser Normalization

The KMO and Bartlett's 2nd Table 17

Kaiser-Meyer-Olkin Measure of Samping

Adequacy 0.941

Bartlett's Test of Approx. Chi-Square 7342

Sphericity df 1801

sig. 0

4.3.2.2 Exploratory factor analysis EFA with satisfaction component As for the measurement scale of job satisfaction, after EFA analysis one factor is deducted with the eigenvalues whose value is 2,611. KMO and Bartlett's test reported its value by 0.737 and significance level priced Sig. = 0.000 (less than

59

0.05). The variables whose value are reported greater than 0.5 (variable SET_1:

0.574; SET_2: 0.832; SET_3: 0.673; SET_4: 0.538; SET_5: 0.565 and SET_6:

0.635), so the observed variables are important components for the satisfaction component.

Rotated Component Matrix (satisfaction component) Table 18

Component

SET_2 .832

SET_3 .673

60

SET_6 .635

SET_1 .574

SET_5 SET_4

.565 .538

Extraction Method: Principal Component Analysis Rotation Method: Varimax with Kaiser Normalization The KMO and Bartlett's (satisfaction component)

Table 19

Kaiser-Meyer-Olkin Measure of Samping

Adequacy 0.737

Bartlett's Test of Approx. Chi-Square 7115

Sphericity df 2611

sig. 0

Thus, the initial research model analysis through results of Cronbach's alpha analysis and exploratory EFA, six proposed components are satisfactory and statistical significance. These components will be used in testing subsequent analyzes.

4.4 RESEARCH MODEL AND HYPOTHESIS TESING

Proposed theoretical model consists of 6 components: (i) Work essence component; (ii) Salary; (iii) Colleague; (iv) Leadership; (v) training and

61

promotion opportunities; (vi) working environment and employee satisfaction on the job. In particular, satisfaction of employees is dependent component, the other six components are independent and are assumed to be the components affecting employee satisfaction.

Regression analysis was conducted to determine the specific weight of each component affecting employee satisfaction. The values of the fators used to run the regression is the average value of the observed variables tested. Regression analysis was performed by means of the overall regression variables with SPSS version 19.0.

4.4.1 Testing regression model between quality components of job and satisfaction

Testing regression model hypothesis between 6 components that are independent variables: (i) Work essence component; (ii) Salary; (iii) Colleague; (iv) Leadership; (v) training and promotion opportunities; (vi) working environment and employee satisfaction on the job is the dependent variable.

The results of the regression model Table 20

Model Summary

Model R R Square Adjusted R

Square

Std. Error of the

Estimate

1 .914 0.863 0.851 0.29176

62

a. Predictors: (Constant), The essence of work, Salary, Colleagues, Leader, Training opportunity and promotion, Work environment

b. Dependent Variable: satisfaction

An R-value is 0.914 showing the relationship between the variables in the model correlating very closely. The report of the regression results showed that the value of R Square by 0.863, which depicted suitability of the model is 86.30%, or in other words, 86.30% of the variation of the variable satisfaction is explained by the six components in the quality of work. Adjusted R Square more accurately reflects the suitability of the model for the overall, the results indicated adjusted R Square whose value is 0.851 (or 85.10%), which means that linear regression model exists between satisfaction and six components in the job.

63

Analysis of variance ANOVA Table 21

Model Sum of

Squares

df Mean

Square

F Sig.

1 Regression Residual Total

141.314 20.124 160.909

5 199 204

28.076 .082

343.112 .000

a. Predictors: (Constant) The essence of work, Salary, Colleagues, Leader, Training opportunity and promotion, Work environment

b. Dependent Variable: Satisfaction

Analysis of variance, ANOVA, showed F value had significant level Sig. = 0.000 , which means the regression model fit the data collected and the variables included are statistical significance at level of 5%. Statistical value of F = 343.112 was used to test the hypothesis H0, here we see a linear relationship is highly significant with p_value <0.05. We can reject the hypothesis H0 that the slope of the six components in the quality of work equal to 0. Thus, the independent variables in the model is related to the dependent variable satisfaction.

64

The statistical parameters of each variable in regression function Table 22

Coefficients(a)

a Dependent Variable: Satisfaction

Model

Unstandardized Standardized

t Sig.

Coefficients Coefficients

B Std.

Error Beta

1 (Constant) -0.186 0.081 -1.904 0.058

The essence of work 0.237 0.03 0.221 7.782 0

Salary 0.247 0.034 0.272 7.166 0

Colleagues 0.234 0.039 0.223 5.676 0

Leader 0.249 0.047 0.239 5.416 0

Training opportunity and

promotion 0.23 0.038 0.232 3.992 0

Work environment 0.202 0.04 0.208 5.316 0

Results of the analysis of the regression coefficients in the model showed that the significance of the components Sig. = 0.000 (less than 0.05). Therefore, we can say that the independent variables have an impact on employee satisfaction on the job.

65

All components of job quality are significant in the model, and effects in the same direction to the satisfaction of employees, so the regression coefficients are positive.

Standardized regression values of the independent variables in the model with values reported respectively: Work essence is 0.221; Salary is 0.272; Colleagues is 0.223; Leadership is 0.239; Training and promotion opportunities is 0.232; Work environment is 0.208. Through the results of regression analysis the model is presented as follow:

SET = 0.237T + 0.247L + 0.234D + 0.249L + 0.23C + 0.202M – 0.186

The model explained 85.10% on the change of R square is due to the independent variables in the model generated, the remaining 14.90% of variation is explained by other variables outside the model.

The model showed that the independent variables are positively influenced the level of user satisfaction at 95% confidence level. Through regression we see, if retained independent variables remaining constant, when assessment of work essence increased by 1 unit, the satisfaction of employees increased by an average of 0.237 units. Similarly, the assessment of salary increased by 1 unit, the employee satisfaction on the job increased at a average of 0.247 units; when evaluating colleagues’s point increased 1 point, the employee satisfaction is increased by a average of 0.234 points; While assessment of leader increased by 1 point, the employee satisfaction for the quality of job increased by a average of 0.249 points;

the assessment of training promotion opportunities increased by 1 point, the employee satisfaction for the quality of job increased a average of 0.23 points; the essessment of working environment increased by 1 point, the employee satisfaction for the quality of job increased a average of 0.202 points. Through the results of

66

Standardized Beta coefficients shows the importance of each independent variable on the dependent variable. Beta values in Table 4:21, tells us that the impact between the six independent variables and the dependent variable, the value of properties in the normative work affecting 22.1% and Satisfaction; standardized regression value of Salaries 27.2% influence to satisfaction; standardized regression value of Colleague affected 22.3% on Satisfaction; standardized regression value of leadership affects 23.9% Satisfaction; standardized regression value of training and promotion opportunities affects 23.2% on Satisfaction;

standardized regression value of working environment impact 20.8% on employee satisfaction in quality of work. Testing regression model with six independent variables and the dependent variable is summarized.

Summary of Hypothesis Testing Table 23

The above table we see that the hypotheses H1, H2, H3, H4, H5 and H6 are acceptable, because the increase of these factors will increase the level of

Hypothesis Testing

results H1: work essence positive relationship with satisfaction accept H2: salary positive relationship with satisfaction accept H3: colleagues positive relationship with satisfaction accept H4: leadership positive relationship with satisfaction accept H5: training and promotion opportunities positive relationship with

satisfaction

accept

H6: working measurement positive relationship with satisfaction accept

67

employee satisfaction in terms of quality of work, or in other words, when employees feel about the quality of work increases, satisfaction also increased.

From the above analysis we can conclude appropriate theoretical models to research data and the research hypothesis is accepted (hypotheses H1, H2, H3, H4, H5 and H6). Results of testing the theoretical model is illustrated by the following figure. Regression coefficient: 0.237

Beta coefficient: 0.221

Regression coefficient: 0.247 Beta coefficient: 0.272

Regression coefficient: 0.234 Beta coefficient: 0.223

Satisfaction

Regression coefficient: 0.230 Beta coefficient: 0.232

Regression coefficient: 0.249 Beta coefficient: 0.239

Regression coefficient: 0.202 Beta coefficient: 0.208

Figure 9 Results of testing the theoretical model

Through Figure 9 shows the importance of the components depends on the absolute value of the regression coefficients that were standardized. The more bigger absolute value Components have, the higher impact on satisfaction levels the components will have. Thus, in Figure 8 we see the satisfaction suffer the most from the salary component (Beta = 0.272); The second important component is leadership (Beta = 0.239); The third important component is training and promotion opportunities (Beta = 0.232); The fourth important component is Colleague (Beta = 0.223); and the another important component is work essence

68

(beta = 0.221), and the final component is working environment component (Beta

= 0.208).

4.4.2 Testing the differences between factors in terms of demograph

This research used the method of one-way analysis of variance to detect the differences between the components according to demographic factors. One factor ANOVA is used to test whether there is a difference existing between research components with respect to demographic factors (according to gender, educations, work position and work unit).

Hypothesis is posed as:

�𝐻0: 𝜇1 = 𝜇2 =, … , = 𝜇𝑘

𝐻𝑎: 𝑛𝑜𝑡 𝑎𝑙𝑙 𝑞𝑢𝑎𝑙.

In which, 𝜇𝑖 is the population mean of i th demographic group.

4.4.2.1 Testing the differences about extent of satisfaction

Testing the differences about extent of satisfaction according to gender, we have the following results.

69

Test of Homogeneity of Variances Table 24

Levene Statistic

df1 df2 Sig.

The essence of work

Significance level Sig. of the independent variables in testing the variances are greater than 0.05. It can be concluded that the variance of the extent of employee satisfaction did not differ according to gender and the results of the ANOVA analysis are used.

70

ANOVA analysis results according to Gender

Through the analysis of variance ANOVA we see the significance level Sig. of all the independent variables are greater than 0.05, it is possible to conclude no differences exists between the independent variables and the extent of satisfaction of employees by gender (or in other words hypothesis H0 is accepted).

71

Similarly, this study uses ANOVA analysis to test the differences in satisfaction levels according to qualifications, positions and departments.

ANOVA analysis results according to Education Table 26 Training opportunity

Between

ANOVA analysis results according to Working position

ANOVA analysis results according to Working department And promotion

Within

For testing differences in satisfaction extent according to qualifications, Location, department, we have the results of testing homogeneity of Variances with a significance level Sig. of the independent variables in testing the variances are greater than 0.05. It can be concluded that the variance of employee satisfaction extent according to qualification, Location, department are not different and the

74

result of ANOVA analysis is not used. In the analysis of variance a factor ANOVA, the value reported on the significance of the six independent variables is greater than 0.05, which means the extent of employee satisfaction according to qualifications, Location, department is not different

Overall assessment on the extent of employee satisfaction with the work of the District 2 Taxation Department is at relatively moderate.

Summary

Chapter 4 presented and solved the problems posed by the thesis. Results of studies with six quality components affecting job satisfaction of employees, including: Work essence; Salary; Colleagues; Leadership; Training and promotion opportunities; Working Environment.

In researching and testing the regression model, the six proposed components are appropriate and meaningful in statistics, a regression model fits the data collected.

In the six components identified in the research model, the impact of the various components on employee satisfaction at work is different. Specifically, the strongest effect to the satisfaction of employee is salary components (Beta = 0.272); the second component is Leadership (Beta = 0.239); The third component is training and promotion opportunities (Beta = 0.232); The fourth component is Colleague (Beta = 0.223; and the fifth component work essence (beta = 0.221), and the final component is working environment (Beta = 0.208). For testing the differences in the extent of satisfaction according to demographic factors, the study also pointed out the differences in satisfaction levels according to gender,

75

qualification, position and department. Specifically, there are no differences in the extent of satisfaction according to gender, qualification, position and department.

The study results of the thesis have contributed to confirm the proposed statements, however, due to the quality of work that is not stable and depends on the perceived level of staff should hinge on the actual conditions of each tax agency should have adjusted the concept and scale accordingly. Besides, the satisfaction of employees on the job depends on many external factors. The next chapter will propose solutions and recommendations to District 2 Taxation department to improve the quality of work in order to satisfy even more the satisfaction of employees.

76

CONCLUSIONS AND CHAPTER 5

RECOMMENDATIONS

5.1 CONCLUSION

The main aim of the study was to determine the impact of the components on employee satisfaction in the work in District 2 Taxation department, construct and evaluate the measurement scale of components. To confirm the effects of these components on the satisfaction of employees, a theoretical model is developed and tested. Theoretical models are built based on theories of service quality, employee satisfaction and the components that have an impact on satisfaction.

Research methods were used to build, measure the scale and test model (presented in Chapter 3), which consisted of two main steps: preliminary research and formal

Research methods were used to build, measure the scale and test model (presented in Chapter 3), which consisted of two main steps: preliminary research and formal

相關文件