Home

Efa Results of the Quality of Working Life Scale


Table 3.4 Survey sample statistics


Information

Result

Proportion

Sex

Male

106

53.00%


Female

94

47.00%

Total

200

100.00%

Title

Manage

13

6.50%


Staff

187

93.50%

Total

200

100.00%

Marital status

Single

148

74.00%

Married

52

26.00%

Total

200

100.00%

Income

< 7 million

124

62%


7 – 12 million

64

32%


>12 million

12

6%

Total

200

100.00%

Maybe you are interested!

Efa Results of the Quality of Working Life Scale

(Source: author's calculation)

3.7 Summary

Chapter III presents the sources of information collected, methods and tools of collection, sample design - sample selection, data processing methods, scales of concepts, characteristics of the survey sample. This is a necessary preparation step for the implementation and determination of research results.


CHAPTER IV

RESEARCH RESULTS

Chapter III presented the research methodology and evaluated the measurement scales of the concepts. Chapter IV presented the results of testing the research model and the proposed hypotheses.

4.1 Measurement model validation

The scale of the quality of work life model is based on the scale of Nguyen et al. (2011). The preliminary survey results showed that there were no differences or changes in the components of the scale for bank employees.

To test the model, the reliability of each component of the quality of work life scale will be assessed through Cronbach's Alpha reliability.

After using Cronbach's Alpha to eliminate variables that do not meet the reliability requirements, the variables that meet the requirements will continue to be included in the exploratory factor analysis (EFA) for the scale of quality of work life, the scale of job satisfaction and the scale of work performance. The purpose of EFA is to explore the structure of the scale of quality of work life and work performance of bank employees in Ho Chi Minh City. Finally, all components are included in the multiple regression analysis to confirm the initial hypothesis.

4.1.1 Preliminary assessment of the scale using Cronbach's Alpha

Cronbach's Alpha coefficient is used to eliminate unsuitable variables first. Variables with corrected item-total correlation coefficients less than 0.30 will be eliminated and the scale selection criteria is when it has a reliability of 0.60 or higher.

According to Hoang Trong and Chu Nguyen Mong Ngoc (2008, page 24): “Many researchers agree that when Cronbach's Alpha is from 0.8 or higher to nearly 1, the measurement scale is good, from nearly 0.7 to nearly 0.8 is usable. Some researchers also suggest that Cronbach's Alpha from 0.6 or higher can be used in the case of


the concept being measured is new or novel to the respondent in the research context (Nunnally, 1978; Peterson, 1994; Slater, 1995)”.

In theory, the higher the Cronbach's Alpha, the better (the scale has high reliability). However, this is not really the case. A Cronbach's Alpha coefficient that is too large (Alpha> 0.95) shows that there are many variables in the scale that are not different from each other (meaning they measure the same content of the research concept). This phenomenon is called redundancy.

The results of Cronbach Alpha reliability testing show that the variables belonging to the component scales all have a reliability greater than 0.50, the total variable correlation of each observed variable is > 0.30. Specifically: the scale for satisfaction of existence needs (TT) has a Cronbach alpha of 0.847; the scale for satisfaction of relationship needs (QH) has a Cronbach alpha of 0.852; the scale for satisfaction of knowledge needs (KT) has a Cronbach alpha of 0.852; the scale for satisfaction of job satisfaction (HL) has a Cronbach alpha of 0.876 and the scale for work performance (KQ) has a Cronbach alpha of 0.836. The total variable correlation coefficients of the scales are all higher than the allowable level (greater than 0.3). Therefore, all observed variables have a reliability level to be used for exploratory factor analysis EFA in the next step.


Table 4.1 Cronbach Alpha test results of the scales



Status


Scale

Number of observed variables

Cronbach's Alpha

Correlation coefficient

between variable-sum minimum

1

Satisfaction of existence needs (TT)

3

0.847

0.705

2

Relationship needs satisfaction (QH)

3

0.852

0.709

3

Satisfaction of knowledge needs (KT)

3

0.852

0.697

4

Job satisfaction (HL)

5

0.876

0.679

5

Work results (KQ)

4

0.836

0.630

(Source: SPSS results)

4.1.2 Exploratory factor analysis (EFA)

All observed variables are put into exploratory factor analysis (EFA) to reduce or summarize the data and calculate the reliability (Sig) of the observed variables to see if they are closely related to each other. When conducting exploratory factor analysis, researchers often pay attention to the following criteria:

- KMO coefficient >= 0.5; significance level of Bartlett test <= 0.05. KMO (Kaiser – Meyer – Olkin measure of sampling adequacy) is an indicator used to consider the appropriateness of EFA, 0.5 ≤ KMO ≤1 then factor analysis is appropriate. Kaiser (1974) suggested that KMO ≥ 0.90 is very good; KMO ≥ 0.80: good; KMO ≥ 0.70: acceptable; KMO ≥ 0.60: acceptable; KMO≥ 0.50: bad; KMO< 0.50: unacceptable.

- Factor loading >= 0.5. According to Hair & colleagues (2006), factor loading is an indicator to ensure the practical significance of EFA. Factor loading > 0.3 is considered to have achieved the minimum level; > 0.4 is considered important; >= 0.5 is considered to have practical significance. Hair & colleagues (2006) also advise that: if choosing the factor loading standard > 0.3, the sample size must be at least 350, if the sample size is about 100, the factor loading standard should be > 0.55, if the sample size is about 50, the factor loading must be > 0.75.

- Total extracted variance >= 50%

- Eigenvalue coefficient >1


- The difference in factor loading coefficients of an observed variable between factors is >= 0.3 to ensure discriminant value between factors.

- Principal Component Analysis extraction method with Varimax rotation and stopping point for extracting factors with eigenvalue >1

4.1.2.1 Quality of work life scale

After testing the scale using Cronbach's Alpha, all 9 observed variables of the 3-component work life quality scale met the requirements and were included in the EFA analysis.

When analyzing EFA with the work quality life scale, the author used the Principal Component Analysis extraction method with Varimax rotation and the stopping point for extracting factors with Eigenvalue >1.

The results of EFA analysis show that 9 observed variables are analyzed into 3 factors. The factor loading coefficients of the observed variables are all > 0.5, so the observed variables are all important in the factors. The difference in factor loading coefficients of an observed variable between factors is all > 0.3, so the discriminant value between factors is guaranteed.

KMO & Bartlett results: KMO coefficient = 0.754 meets the requirement > 0.5 so EFA is suitable for the data. The Chi-Square statistic of Bartlett test reaches 1.131 with significance level Sig = 0.000; therefore, the observed variables are correlated with each other in the overall scope.

Eigenvalue = 1.011 >1 meets the requirements, the stopping point is at the 3rd factor with the extracted variance reaching 77.218%, meaning that the 3 extracted factors explain 77.218% of the data variation (see Appendix 6).


Table 4.2 EFA results of the quality of working life scale


STT

Variable name

Factor name

1

2

3


1

KT3

0.849



Satisfaction of knowledge needs (KT)

2

KT1

0.822



3

KT2

0.787



4

QH1


0.848


Relationship needs satisfaction (QH)

6

QH2


0.842


7

QH 3


0.810


8

TT2



0.834


Satisfaction of existence needs (TT)

9

TT1



0.827

10

TT3



0.788

(Source: SPSS results)

The first factor includes the following three observed variables:

KT1: My job allows me to perform to the best of my ability KT2: My job helps me improve my professional skills KT3: My job helps me develop my creativity

This factor is named Satisfaction of knowledge needs , symbolized as KT.

The second factor consists of 3 observed variables: QH1: I have good friends at the bank

QH2: After work, I have enough time to relax and entertain. QH3: I am respected at the bank.

This factor is named Relationship Need Satisfaction , symbolized as QH.

The third factor consists of 3 observed variables: TT1: The bank provides me with good welfare regime TT2: I am satisfied with my income at the bank

TT3: My current job at the bank ensures my livelihood. This factor is named Existence Need Satisfaction , denoted by TT.

4.1.2.2 Job Satisfaction Scale

The results of factor analysis for the job satisfaction scale showed that 1 factor was extracted and no observed variables were eliminated. With coefficient


KMO = 0.866, Chi-Square test = 476.243, significance level Sig = 0. Factor loading coefficients of all variables are above 0.7; extracted variance is 67.080%. Thus, all observed variables of the job satisfaction scale meet the requirements (see Appendix 7).

4.1.2.3 Performance Measurement Scale

Similarly, the factor analysis results for the work performance scale showed that there was also 1 factor extracted and no observed variables were eliminated. With the KMO coefficient = 0.810, Chi-Square test = 301.927, significance level Sig = 0. The factor loading coefficient of all variables was above 0.7; the extracted variance was 67.378%. Thus, all observed variables of the work performance scale met the requirements (see Appendix 8).

Thus, after performing exploratory factor analysis (EFA), officially testing the reliability of the work quality life scale, there was no variable elimination, so the research model remained the same as the original.

4.2 Regression analysis

Model analysis: includes 2 regression models: (1) analysis of the impact of quality of work life on work performance, (2) quality of work life on job satisfaction.

Issues to consider in regression models:

- First, before performing regression, we consider the linear correlation between all variables (independent variables and dependent variables, and between independent variables with each other), to see the degree of close relationship between variables.

- Second, test the suitability of the regression model to the data set using the adjusted coefficient of determination (adjusted R2 ) , which measures the percentage of variation explained in the dependent variable taking into account the relationship between sample size and the number of independent variables in the multiple regression model, thus avoiding exaggerating the model's ability to explain the dependent variable; test the suitability of the overall model using the F statistic.

- Third, test the significance level of the partial coefficients using the t statistic.


- Fourth, check for violations of assumptions (linear relationship assumption, residual assumptions: constant variance, normal distribution, independence and assumption of no correlation between independent variables), because if the assumptions are violated, the estimation results will no longer be reliable.

- Fifth, determine the importance of variables in the model.

4.2.1 Correlation analysis

Before conducting regression analysis, we will consider the linear correlation between the dependent variable and each independent variable, as well as between the independent variables with each other. The larger the correlation coefficient between the dependent variable and the independent variables, the higher the relationship between the dependent variable and the independent variables, and thus regression analysis may be appropriate. On the other hand, if there is a large correlation between the independent variables, this means that multicollinearity may occur in the regression model.

The Person correlation coefficient is used to examine the linear correlation between the dependent variable and each independent variable, as well as between independent variables with each other. This coefficient is always in the range from -1 to 1, taking the absolute value, if it is greater than 0.6, we can conclude that the relationship is tight, and the closer it is to 1, the tighter the relationship, if it is less than 0.3, we know that the relationship is loose.

Table 4.3 Correlation coefficient

(N=200)


Correlate

HL

Result

TT

QH

KT

HL

1

0.986

0.748

0.715

0.735

Result

0.986

1

0.766

0.735

0.742

TT

0.748

0.766

1

0.478

0.568

QH

0.715

0.735

0.478

1

0.483

KT

0.735

0.742

0.568

0.483

1

(Source: SPSS results)

The analysis results show that there is a correlation between Job satisfaction and the independent variables Existence needs satisfaction, Relationship needs satisfaction, Knowledge needs satisfaction and this relationship is relatively strong.

Comment


Agree Privacy Policy *