2.5.2.2 Cronbach's Alpha test for dependent variable
Table 2.6 : Cronbach's values for dependent variables
Observation variable
Average scale if type variable | Scale variance if variable is excluded | Total variable correlation | Cronbach's Alpha if this variable is eliminated | |
Dependent variable scale Work motivation; Cronbach's Alpha = 0.797 | ||||
DLLV1 | 7.98 | 1,085 | 0.622 | 0.746 |
DLLV2 | 7.92 | 1,101 | 0.667 | 0.697 |
DLLV3 | 7.98 | 1,155 | 0.637 | 0.729 |
Maybe you are interested!
-
Results of Testing Cronbach's Alpha Coefficient of Independent Variable -
Cronbach'S Alpha Coefficient Test Results of the Ability Scale -
Cronbach'S Alpha Test Results of Service Competency Round 1 -
Measuring Cronbach'S Alpha Coefficients of Dependent Variables -
Cronbach'S Alpha Reliability Test
(Source: Synthesized from analysis results on SPSS)
It is found that the Cronbach's Alpha coefficient for the factor "Work motivation" is 0.797, which is in the high correlation range. The observed variables all have a total correlation coefficient greater than 0.3 and have a Cronbach's Alpha coefficient if the variable is eliminated less than the total Cronbach's Alpha, so all 3 observed variables are retained. From the results, it can be seen that this scale is reliable enough for further analysis.
2.5.3 Exploratory factor analysis EFA
Exploratory factor analysis is used to examine the relationships between variables in the research data and the ultimate goal is to summarize and reduce observed variables into the same factor, ensuring discriminant and convergent validity for the scales.
Exploratory factor analysis EFA needs to achieve the following analytical steps:
+ EFA (KMO) suitability test
The KMO index must meet the following condition: 0.5 < KMO <1. When this condition is satisfied, factor analysis is suitable for analyzing real data.
+ Correlation test of observed variables in representative measure (Bartlett):
Bartlett index has Sig. < 0.05, meaning that the observed variables have linear correlation with the representative factor.
+ Test the level of explanation of observed variables for factors:
Through two variance indices extracted Eigenvalues > 1, Cumulative index > 50%. Level.
the explanatory power of the observed variables for the factor.
2.5.3.1 Results of exploratory factor analysis EFA of independent variables
Table 2.6: KMO and Bartlett tests for independent variables
TT
PARAMETER | Value | Condition | Comment | |
1 | KMO | 0.779 | ≥0.5 | Meet the requirements |
2 | Sig, of Bartlett's Test | 0.000 | ≤0.05 | Meet the requirements |
3 | Eigenvalues | 1,292 | >1 | Meet the requirements |
4 | Total extracted variance | 69.371% | ≥50% | Meet the requirements |
(Source: Synthesized from analysis results on SPSS) KMO index = 0.779 > 0.5: Suitable for analyzing real data.
Bartlett index has Sig. = 0 < 0.05, observed variables have linear correlation with representative factor.
Eigenvalues = 1.292 > 1, representing the portion of variation explained by each factor, then the extracted factor has the best summary of information.
Cumulative shows that the extracted variance value is 69.371%. The factor variation explains 69.371% of the variation in the observed variable.
Table 2.7 : Factor rotation matrix
1 | 2 | 3 | 4 | 5 | 6 | |
BCCV4 | 0.829 | |||||
BCCV2 | 0.827 | |||||
BCCV1 | 0.750 | |||||
BCCV3 | 0.730 | |||||
DKLV4 | 0.887 | |||||
DKLV3 | 0.886 | |||||
DKLV2 | 0.765 | |||||
DKLV1 | 0.599 | |||||
LTPL1 | 0.784 | |||||
LTPL2 | 0.779 | |||||
LTPL4 | 0.730 | |||||
LTPL3 | 0.704 | |||||
PCLD1 | 0.829 | |||||
PCLD3 | 0.807 | |||||
PCLD4 | 0.775 | |||||
PCLD2 | 0.737 | |||||
QDN4 | 0.869 | |||||
QDN2 | 0.815 | |||||
QDN1 | 0.703 | |||||
QDN3 | 0.690 | |||||
DTTT3 | 0.838 | |||||
DTTT2 | 0.745 | |||||
DTTT1 | 0.733 |
(Source: Synthesized from analysis results on SPSS)
Based on the matrix results table above, we see that all factor loading coefficients of all variables are greater than 0.5. After performing rotation, the factors have no disturbance between observed variables, showing that observed variables in the same factor all have
convergent validity and eigenvalue. Thus, based on the results of the factor rotation matrix, it shows that all 23 initial observed variables through factor analysis have extracted 6 factors as the initial research model.
2.5.3.2 Results of exploratory factor analysis EFA of dependent variable
Conducting exploratory factor analysis (EFA) on the dependent variable "Work motivation" we obtained the following results:
Table 2.8 : KMO and Bartlett tests for dependent variables
TT
PARAMETER | Value | Condition | Comment | |
1 | KMO | 0.708 | ≥0.5 | Meet the requirements |
2 | Sig, of Bartlett's Test | 0.000 | ≤0.05 | Meet the requirements |
3 | Eigenvalues | 2,138 | >1 | Meet the requirements |
4 | Total extracted variance | 71,267% | ≥50% | Meet the requirements |
(Source: Synthesized from analysis results on SPSS)
KMO index = 0.708 > 0.5: Suitable for analyzing real data.
Bartlett index has Sig. = 0 < 0.05, observed variables have linear correlation with representative factor.
Eigenvalues = 2.138 > 1, representing the portion of variation explained by each factor, then the extracted factor has the best summary of information.
Cumulative shows that the extracted variance value is 71.267%. The factor variation explains 71.267% of the variation in the observed variable.
2.5.4 Determining the level of influence of factors affecting employees' work motivation using correlation regression method
After EFA, 6 factors were included in the model testing. The value of each factor is the average value of the observed variables belonging to that factor. Correlation analysis was conducted between the 6 factors found after EFA analysis with the dependent variable "Work motivation" before conducting multiple regression.
variables to test the impact level of factors on "Work motivation" such as
how
2.5.5 Building a regression model
In the regression analysis model, the dependent variable is “Work motivation”. The independent variables are the factors extracted from the observed variables from the EFA factor analysis. The regression model is as follows:
DLLV= β0 + β1*BCCV + β2*DKLV + β3*LTPL + β4*PCLD + β5*QHDN +
β6*DTTT + ei
In there:
DLLV: Value of dependent variable
BCCV: Value of the independent variable “Nature of work” DKLV: Value of the independent variable “Working conditions”
LTPL: Value of independent variable “Salary, bonus, welfare policy” PCLD: Value of independent variable “Leadership style”
QHDN: Value of independent variable “Relationship with colleagues” DTTT: Value of independent variable “Training and promotion”
ei : Factors other than the independent variable. The hypotheses are set out as follows:
H0: The factors have no correlation with the dependent variable "Work motivation" of employees at Phu Hoa An Textile Joint Stock Company.
H1: The factor “BCCV” is correlated with the work motivation of employees. H2: The factor “DKLV” is correlated with the work motivation of employees. H3: The factor “LTPL” is correlated with the work motivation of employees. H4: The factor “PCLD” is correlated with the work motivation of employees. H5: The factor “QHDN” is correlated with the work motivation of employees. H6: The factor “DTTT” is correlated with the work motivation of employees.
2.5.6 Correlation coefficient test
Table 2.9 : Pearson correlation analysis
Factor
Common motivation | ||
BCCV | Pearson Correlation Sig. (2-tailed) | 0.301 0.001 |
DKLV | Pearson Correlation Sig. (2-tailed) | 0.375 0.000 |
LTPL | Pearson Correlation Sig. (2-tailed) | 0.386 0.000 |
PCLD | Pearson Correlation Sig. (2-tailed) | 0.312 0.000 |
QDN | Pearson Correlation Sig. (2-tailed) | 0.184 0.036 |
DTTT | Pearson Correlation Sig. (2-tailed) | 0.375 0.000 |
(Source: Synthesized from analysis results on SPSS)
The above results show that the independent variables have a fairly high correlation with the independent variable work motivation, the correlation level ranges from 0.184 to 0.386. In addition, the independent variables also have a fairly strong correlation with each other, so it is very easy for multicollinearity to occur, making the estimated model no longer good. Therefore, to consider this phenomenon, the author conducted a multicollinearity test in the regression step of the model.
2.5.7 Regression analysis
Multiple regression analysis is to estimate the influence of factors on the dependent variable and to test the hypotheses proposed in the research model.
The method of performing regression on SPSS is to enter factors at the same time (Enter). To evaluate the level of explanation of factors on the dependent variable through the adjusted R-square coefficient (Adjusted R-square) to evaluate the suitability of the model and the level of
explanation of the variables because it does not inflate the model's fit (Hoang Trong and Chu Nguyen Mong Ngoc, 2008).
The standardized Beta coefficient is used to evaluate the importance of each factor. The higher the standardized Beta coefficient of a variable, the greater the impact of that variable on customer satisfaction with service quality (Hoang Trong and Chu Nguyen Mong Ngoc, 2008).
Because the value of the unstandardized regression coefficient (B) depends on the scale, we cannot use it to compare the impact of independent variables on the dependent variable in the same model. The standardized regression coefficient (beta, symbol β) is the coefficient that we have standardized the variables. Therefore, they are used to compare the impact of dependent variables on independent variables. The independent variable with a larger weight means that the variable has a strong impact on the dependent variable. Moreover, the unstandardized regression coefficient explains the change in a dependent variable based on a change in an independent variable under the condition that other independent variables remain unchanged. The explanations are more mathematical, so the author uses the standardized regression coefficient to solve the impact of factors in the research model.
The F test is used to test the suitability of the model with the original data set. If the significance level of the test is <0.05, it can be concluded that the regression model is suitable for the data set (Hoang Trong and Chu Nguyen Mong Ngoc, 2008).
The regression results using the Enter method are presented in the following tables:
2.5.7.1 Multiple Linear Regression
Table 2.1 0 : Regression results and multicollinearity
Model
Unstandardized coefficient chemical | Coefficient standardize | T | Sig. | Multicollinearity statistics line | ||||
B | Std. Error | Beta | Tolerance | VIF | ||||
(Constant) | -2,574E-016 | 0.053 | 0.000 | |||||
LTPL | 0.386 | 0.053 | 0.386 | 7,260 | 0.000 | 1,000 | 1,000 | |
DKLV | 0.375 | 0.053 | 0.375 | 7,053 | 0.000 | 1,000 | 1,000 | |
DTTT | 0.375 | 0.053 | 0.375 | 7,038 | 0.000 | 1,000 | 1,000 | |
PCLD | 0.312 | 0.053 | 0.312 | 5,856 | 0.000 | 1,000 | 1,000 | |
BCCV | 0.301 | 0.053 | 0.301 | 5,652 | 0.000 | 1,000 | 1,000 | |
QDN | 0.184 | 0.053 | 0.184 | 3,458 | 0.001 | 1,000 | 1,000 | |
(Source: Synthesized from analysis results on SPSS)
In the Sig value column, we see 6 factors: LTPL, DKLV, DTTT, PCLD, BCCV,
QHDN all have sig. < 0.1 meaning that the variables are statistically significant at the 10% level.
The table above shows that the Tolerance values are all > 0.1 and the VIF coefficients are all <3. Conclusion: There is no multicollinearity in the model.
2.5.7.2 Conformance testing
In the multiple regression analysis results, adjusted R² 
= 0.635, the factors LTPL, DKLV, DTTT, PCLD, BCCV,
QHDN can explain 63.5% of the variation in the DLLV variable (Work motivation).
Durbin-Watson coefficient (d) = 1.829, ranging from 1-3. The model has no autocorrelation between residuals in the model, the model has practical significance.