OC 4
For me, this is one of the best organizations to work for. | |
OC 5 | I am really concerned about the state of the organization. |
Maybe you are interested!
-
Testing of Scales – Cronbach'S Alpha Reliability Coefficient. -
Testing the Reliability of the Scale Using Cronbach'S Alpha Coefficient -
Testing the Reliability of the Scale Through Cronbach'S Alpha Coefficient -
Results of Testing Cronbach's Alpha Coefficient of Independent Variable -
Cronbach's Alpha Reliability Coefficient of Components of the Scale of Factors Affecting Investment Capital Attraction for Tourism in Ba Ria Vung Tau Province
(Source: Mowday, Steers, and Porter, 1979)
3.5. Data analysis method
3.5.1. Data analysis method
Step 1: Conduct a survey to collect opinions from bank employees in Ho Chi Minh City. The survey was conducted within 30 minutes (the survey forms were distributed and collected during the direct meeting with bank employees). 250 survey forms were distributed, 233 were collected. Of which, 23 were invalid for reasons such as choosing a question with no options or more than 2 options. After that, the information was further filtered, and 187 qualified survey forms were included in the analysis. The data was analyzed by SPSS version 22 software.
Step 2: Conduct descriptive statistics of survey subjects by gender, age, education level, position, and seniority.
Step 3: Verify and calibrate the scale by verifying the reliability of the scale using Cronbach's Alpha coefficient and exploratory factor analysis (EFA). After performing Cronbach's Alpha coefficient verification, the questions with the largest coefficient and the closest relationship to the factors to be measured will be selected.
Step 4: Regression testing for each hypothesis, from which accept or reject the research hypotheses, including:
- Test the correlation between factors;
- Multivariate regression testing between independent variables affecting the factor of Organizational Commitment.
3.5.2. Reliability testing using Cronbach's Alpha coefficient
According to Hoang Trong and Chu Nguyen Mong Ngoc (2008), Cronbach's Alpha is a statistical test of the degree of correlation between variables in a scale. According to Nunnally and Berstein (1994), if Cronbach's Alpha > 0.6 and the Item-Total Correlations ≥ 0.3, the scale is acceptable in terms of reliability. In theory, the higher the Cronbach's Alpha, the better (the more reliable the scale). However, if the Cronbach's Alpha coefficient is too large (> 0.95), it shows that many variables in the scale are not different from each other (Nguyen Dinh Tho, 2011). After analyzing the reliability test, the author retains the questions (observed variables) with Cronbach's Alpha coefficient > 0.6 and correlation with the total variable ≥ 0.3.
3.5.3. Exploratory factor analysis
Exploratory factor analysis (EFA) is used to reduce and summarize variables into factors. Factor analysis helps to determine the relationship between multiple variables and find out the representative factors for observed variables. In this study, exploratory factor analysis was conducted for the independent variables: work environment, organizational learning culture, quality of communication between employees and leaders, employee engagement and the dependent variable: organizational commitment. The stages of an exploratory factor analysis EFA include:
- First, it is necessary to conduct the Kaiser – Meyer – Olkin (KMO) test and the Barlett test with the hypothesis H0: there is no relationship between the observed variables. According to Hoang Trong and Chu Nguyen Mong Ngoc (2008), the Barlett test is used to examine the hypothesis that the variables have no overall correlation, while the KMO test is used to examine the appropriateness of EFA. The KMO coefficient > 0.5 and the Sig of the Barlett's test
< 0.05 shows that the relationship between observed variables is large enough, suitable for conducting exploratory factor analysis EFA (Hai et al., 1998).
- Next, we determine the number of factors based on percentage.
explained variation and Eigenvalue. Factor analysis results are accepted when the total variance extracted is > 50% and the Eigenvalue is greater (Gerbing and Anderson, 1998). In the social sciences, the required percentage of explained variation is about 50-60% (Hair et al., 1998).
Factor Loading is an indicator to ensure the practical significance of EFA analysis. According to Hair et al. (1998), Factor Loading > 0.3 is considered the minimum level, Factor Loading > 0.4 is considered important, Factor Loading ≥ 0.5 is considered to have practical significance. In this study, the author chose the criterion Factor Loading ≥ 0.5.
3.5.4. Building a regression program and testing research hypotheses
From the results of the exploratory factor analysis, the author built a multivariate linear equation and conducted regression testing for each research hypothesis. In addition, the study also tested the differences in organizational commitment between demographic groups such as: gender, age, education level and seniority, and job position.
Summary:
In this chapter, the author presents the research design and proposes the sampling method and the method of determining the sample size. The response options are on a Likert scale with 5 levels from: (1) completely disagree, (2) disagree, (3) no opinion, (4) agree, (5) completely agree. The author uses SPSS version 22 software to analyze the data: descriptive statistics of the survey subjects; verify and calibrate the scale by verifying the reliability of the scale using Cronbach's Alpha coefficient and exploratory factor analysis (EFA); regression testing for each hypothesis, thereby accepting or rejecting the research hypotheses.
CHAPTER 4 RESEARCH RESULTS
The research method used by the author in the research process has been presented specifically in chapter 3. Next, in chapter 4, the author will present the results of the analysis of survey sample information, scale testing and the results of testing the hypotheses proposed in chapter 2.
4.1. Qualitative research results
After conducting interviews with experts using face-to-face interview methods, it was found that: Working environment, learning culture, quality of leadership-employee communication and employee engagement have a positive impact on organizational commitment. This is an important basis for continuing research.
4.2. Descriptive statistics
From the observed variables, the author conducted statistics according to demographic characteristics: gender, age, education level, seniority, position. The results are as follows:
Table 4.1: Demographic characteristics of the study sample
Demographic characteristics
Quantity | Proportion | |
Sex | ||
Male | 85 | 45.5 |
Female | 102 | 54.5 |
Age | ||
Under 30 years old | 87 | 46.5 |
30 to 40 years old | 79 | 42.2 |
Over 40 years old | 21 | 11.2 |
Education level | ||
Secondary/College | 17 | 9.1 |
University | 89 | 47.6 |
Postgraduate | 81 | 43.3 |
Seniority | ||
Under 2 years | 33 | 17.6 |
From 2-5 years | 107 | 57.2 |
Over 5 years | 47 | 25.1 |
Location | ||
Staff/Specialist | 121 | 64.7 |
Controller | 41 | 21.9 |
Team Leader/Chief department (deputy department) | 25 | 13.4 |
(Source: Author's survey 2018) | ||
Regarding gender, in 187 observations, there were 102 female employees (54.5%) and 85 male employees (45.5%).
In terms of age, there are 87 employees under 30 years old (accounting for 46.5%); 79 employees between 30 and 40 years old (accounting for 42.2%) and 21 employees over 40 years old (accounting for 11.2%).
Regarding education level, there are 17 employees with Intermediate/College degree (accounting for 9.1%), 89 employees with University degree (accounting for 47.6%), and 81 employees with Postgraduate degree (accounting for 43.3%).
In terms of seniority, there are 33 employees with less than 2 years of seniority (accounting for 17.6%), 107 employees with 2 to 5 years of seniority (accounting for 57.2%), and 47 employees with more than 5 years of seniority (accounting for 25.1%).
In terms of positions, there are 121 employees/specialists (accounting for 64.7%), 41 controllers (accounting for 21.9%) and 25 Team Leaders/Deputy Department Heads (Deputy Department Heads) (accounting for 13.4%).
Table 4.2: Pairwise data statistics
Sex | Age | |||||
Male | Female | Under 30 year old | From 30 to 40 years old | Over 40 year old | ||
Seniority | Under 2 years | 33 | 0 | 33 | 0 | 0 |
From 2-5 years | 52 | 55 | 54 | 53 | 0 | |
Over 5 years | 0 | 47 | 0 | 26 | 21 | |
Education level | Intermediate/Advanced class | 17 | 0 | 17 | 0 | 0 |
University | 68 | 21 | 70 | 19 | 0 | |
Postgraduate | 0 | 81 | 0 | 60 | 21 | |
(Source: Author's survey 2018)
4.3. Cronbach's Alpha reliability test of the scale
The scale of working environment, the scale of learning culture in the organization, the scale of the quality of exchange between leaders and employees, the scale of employee engagement are scales with different measurement contents. Therefore, before using these scales for measurement, it is necessary to test the level of rigor and eliminate the observed variables that do not meet the requirements.
4.3.1. Testing the reliability of the work environment scale in the organization
The work environment scale consists of 5 questions, corresponding to the observed variables: WE1, WE2, WE3, WE4, WE5. The results of Cronbach's Alpha reliability test of the scale are as follows:
Table 4.3: Cronbach's Alpha coefficient of work environment scale
(first inspection)
Observation variable
Average scale if type variable | Scale variance if variable is excluded | Correlation with total variable | Cronbach's alpha if variable is excluded | |
Cronbach's Alpha reliability = 0.776 | ||||
WE1 | 14.99 | 12,640 | 0.680 | 0.689 |
WE2 | 14.25 | 12,555 | 0.608 | 0.714 |
WE3 | 14.05 | 12,831 | 0.617 | 0.711 |
WE4 | 14.99 | 12,398 | 0.788 | 0.656 |
WE5 | 14.06 | 17,389 | 0.129 | 0.857 |
(Source: Author's processing results 2018)
The Cronbach's Alpha coefficient of the work environment scale is: 0.776. In Nunally's (1978) study, a good measurement scale will have a Cronbach's Alpha coefficient > 0.8 (cited by Hoang Trong and Chu Nguyen Mong Ngoc, 2008). In addition, the correlation coefficient with the total variable is > 0.3 (higher than the limit). However, the component variable WE5 has a correlation with the total variable of 0.129 (< 0.3), so we remove this variable and test the reliability of the scale with the variables WE1, WE2, WE3, WE4, the results are as follows:
Table 4.4: Cronbach's Alpha coefficient of work environment scale
(2nd inspection)
Observation variable
Average scale if type variable | Scale variance if variable is excluded | Correlation with total variable | Cronbach's alpha if variable is excluded | |
Cronbach's Alpha reliability = 0.857 | ||||
WE1 | 10.97 | 10,504 | 0.677 | 0.828 |
WE2 | 10.22 | 9,982 | 0.670 | 0.834 |
WE3 | 10.02 | 10,150 | 0.694 | 0.822 |
WE4 | 10.96 | 10,348 | 0.777 | 0.791 |
(Source: Author's processing results 2018)
The Cronbach's Alpha coefficient of the work environment scale after the second test is: 0.857. In Nunally's (1978) study, a good measurement scale will have a Cronbach's Alpha coefficient > 0.8 (cited by Hoang Trong and Chu Nguyen Mong Ngoc, 2008). In addition, the correlation coefficient with the total variable is > 0.3 (higher than the limit). Therefore, the observed variables WE1, WE2, WE3, WE4 are used in the following analyses.
4.3.2. Testing the reliability of the scale measuring learning culture in organizations
The work learning culture scale consists of 5 questions, corresponding to the observed variables: OLC1, OLC2, OLC3, OLC4, OLC5. The results of Cronbach's Alpha reliability test of the scale are as follows:
Table 4.5: Cronbach's Alpha coefficient of learning culture scale
Observation variable
Average scale if type variable | Scale variance if variable is excluded | Correlation with total variable | Cronbach's alpha if variable is excluded | |
Cronbach's Alpha reliability = 0.892 | ||||
OLC1 | 15.01 | 7,403 | 0.687 | 0.879 |
OLC2 | 15.09 | 6,939 | 0.765 | 0.862 |
OLC3 | 15.31 | 6,871 | 0.772 | 0.860 |
OLC4 | 15.10 | 7,206 | 0.744 | 0.866 |
OLC5 | 15.10 | 7,539 | 0.713 | 0.874 |
(Source: Author's processing results 2018)
The Cronbach's Alpha coefficient of the learning culture scale is: 0.892. In Nunally's (1978) study, a good measurement scale will have a Cronbach's Alpha coefficient > 0.8 (cited by Hoang Trong and Chu Nguyen Mong Ngoc, 2008). In addition, the correlation coefficient with the total variable is > 0.3 (higher than the limit). Therefore, the 5 observed variables OLC1, OLC2, OLC3, OLC4, OLC5 are kept intact and used in the following analyses.
4.3.3. Reliability testing of the scale measuring the quality of leader-employee communication in the organization
The scale measuring the quality of leader-employee communication in the organization consists of 7 questions (the measurement variable was proposed by Scandura and Graen (1984). The reliability of the scale is 0.87, which has been tested by (Joo, 2010). The author has tested the above variables and obtained the following results:
Table 4.6: Cronbach's Alpha coefficient of the scale measuring the quality of communication between
leader-employee
Observation variable
Medium scale if variable type | Variance scale if variable type | Correlation with total variable | Cronbach's alpha if variable type | |
Cronbach's Alpha reliability = 0.891 | ||||
LMX1 | 20.79 | 34,897 | 0.769 | 0.865 |
LMX2 | 20.99 | 36,124 | 0.676 | 0.877 |
LMX3 | 21.20 | 36,988 | 0.635 | 0.881 |
LMX4 | 21.05 | 34,642 | 0.717 | 0.872 |
LMX5 | 20.82 | 36,479 | 0.650 | 0.880 |
LMX6 | 20.93 | 37,640 | 0.624 | 0.882 |
LMX7 | 20.84 | 34,906 | 0.739 | 0.869 |
(Source: Author's processing results 2018)
The Cronbach's Alpha coefficient of the leader-employee communication quality scale is: 0.891. In Nunally's (1978) study, a good measurement scale will have a Cronbach's Alpha coefficient > 0.8 (cited by Hoang Trong and Chu Nguyen Mong Ngoc, 2008). In addition, the correlation coefficient with the total variable is > 0.3 (higher than the limit). Therefore, the 7 observed variables LMX1, LMX2, LMX3, LMX4, LMX5, LMX6, LMX7 are kept intact and used in the following analyses.
4.3.4. Reliability testing of the scale measuring employee commitment to the organization
To measure the employee engagement factor with the organization, the author used the proposed scale consisting of 6 corresponding questions: EE1, EE2, EE3, EE4, EE5, EE6. The results of the reliability test of the employee engagement scale are as follows:
Table 4.7: Cronbach's Alpha coefficient of employee engagement scale
(1st inspection)
Observation variable
Average scale if type variable | Scale variance if variable is excluded | Correlation with total variable | Cronbach's alpha if variable is excluded | |
Cronbach's Alpha reliability = 0.834 | ||||
EE1 | 17.81 | 20,820 | 0.692 | 0.789 |
EE2 | 17.17 | 19,723 | 0.719 | 0.782 |
EE3 | 16.82 | 21,741 | 0.679 | 0.792 |
EE4 | 17.50 | 23,004 | 0.659 | 0.799 |
EE5 | 17.68 | 21,916 | 0.736 | 0.783 |
EE6 | 17.37 | 26,460 | 0.229 | 0.878 |
(Source: Author's processing results 2018)
The Cronbach's Alpha coefficient of the employee engagement scale is: 0.834 (> 0.6). In Nunally's (1978) study, a good measurement scale will have a Cronbach's Alpha coefficient > 0.8 (cited by Hoang Trong and Chu Nguyen Mong Ngoc, 2008). In addition, the correlation coefficient with the total variable is > 0.3 (higher than the limit). However, the component variable EE6 has a correlation with the total:
0.229 < 0.3 so we remove this variable and test the reliability of the scale with variables EE1, EE2, EE3, EE4, EE5, the results are as follows:
Table 4.8: Cronbach's Alpha coefficient of employee engagement scale
(2nd inspection)
Observation variable
Average scale if type variable | Scale variance if variable is excluded | Correlation with total variable | Cronbach's alpha if variable is excluded | |
Cronbach's Alpha reliability = 0.878 | ||||
EE1 | 14.31 | 16,785 | 0.720 | 0.850 |
EE2 | 13.67 | 16,072 | 0.715 | 0.854 |
EE3 | 13.32 | 17,744 | 0.695 | 0.856 |
EE4 | 14.00 | 18,720 | 0.698 | 0.857 |
EE5 | 14.18 | 17,924 | 0.752 | 0.844 |
(Source: Author's processing results 2018)
The Cronbach's Alpha coefficient of the employee engagement scale after the second test is: 0.878. In Nunally's (1978) study, a good measurement scale will have a Cronbach's Alpha coefficient > 0.8 (cited by Hoang Trong and Chu Nguyen Mong Ngoc, 2008). In addition, the correlation coefficient with the total variable is > 0.3 (higher than the limit). Therefore, the observed variables EE1, EE2, EE3, EE4, EE5 are used in the following analyses.
4.3.5. Reliability testing of organizational commitment scale
The results of Cronbach's Alpha test (testing the reliability of the organizational commitment scale) are presented in Table 4.9.