Chapter Overview
This chapter gives detailed descriptive and empirical results. First, the descriptive results are presented in a table form for all the variables and then they are explained in detail in terms of the overall means for the findings. A table of frequencies for demographics is also presented. Later the tables of empirical results for the hypotheses are presented based on the model as 4.1, 4.2 and 4.3 in that order. For each of the table presented, the meanings of the results are presented below. In this way each item tested is treated as an individual hypothesis. A thorough discussion follows after the last part (4.3). In the discussion part, the eight hypotheses are explained in details in terms of what the results mean.
52
53 Descriptive results discussion
Based on the descriptive results in Table 4.1., several assumptions and conclusions can be made. To have a fair and unbiased judgement, cut-points for the means have to be set in order to judge the perceptions of the employees as either being too low, low, high or very high. Since the questionnaire used the Likert type scaling where 1 is very low and 5 is very high, the researcher takes a cut point of 2.5 to be very low and a cut-point of 3.5 to be good or satisfactory and 4 to be high. For standard deviations and variance, 2.5 will be considered to be high. Another consideration to this is the total number of valid responses to the individual variable items. The lowest number of valid responses so far is 111 while the highest figure is 116. As such, most of the numbers fall within this range. These rules are applicable to all the other variable items that are discussed in this section.
Based on the above definitions, the means of the salary are all low-below 3.5 cut-point of good. The values range from (2.12 to 3.32), with none other reaching 3.5. We can say that on average employees have low perceptions of their salary. This means that the employees are not satisfied with their salaries that the employer gives them. It is also proper to conclude that the majority of the employees have a similar feeling or perception based on the standard deviations and the variance observed on these items. The standard deviations are not too big (0.956 to 1.168) and so are the variance figures (0.915 to 1.364). This means that employees have similar levels of satisfaction (lack of satisfaction in this case).
Looking at the means for job enrichment, the results are a bit higher than those of salary although they are not too high. They range from 2.69 to 3.67. This means that the employees have better perception of their job in terms of how they are aligned or enriched.
Most of the means are above 2.5 but are below 3.5. This means that their perception is still low but higher than the perception for salary. Although this is the case, the values are not good enough to say the employees have well aligned and enriched jobs that may boost their satisfaction. The standard deviations are not too varied, (0.995 to 1.106) and so are the variances (0.991 to 1.223). In practice this means that a large number of employees have similar levels of perception and their differences are not too far apart. The variances show that there are no outliers and the interpretation of the results above holds.
54
The means for job stability/security are equally low but higher than those for job enrichment. They range from 2.77 to 3.85. The standard deviations range from 1.038 to 1.183 and their variances range from 1.077 to 1.413. This means that most of the employees hold a similar perception of the security that employers accord their jobs. The values are still low considering that job stability or security is a very important factor in trying to retain employees in every organization. Low perception of job security may have serious negative implications in terms of employee satisfaction and commitment for the job. If employees highly feel that their jobs are not really secure, they are most likely to find other jobs that would provide them with security and also satisfaction.
The values for job satisfaction are also low although they are not actually below 2.5.
They range from 3.14 to 3.27 and so their standard deviations (0.949 to 1.205) and variances are also within the same range (0.900 to 1.452). In essence this means that employees do not really feel satisfied with their jobs. The lack of satisfaction however is not too low and neither is their satisfaction too high. This low satisfaction may be projected from the previous low means of the independent variables of salary, job enrichment and job security.
A cut point of 3.5 would be deemed satisfactory enough to say the job satisfaction is relatively high. Therefore there is low satisfaction among employees.
The values of employee commitment are not any better-they are all low too. They range between 1.91 which is very low to 3.02 which is a bit better but not high. This means that on overall there is low commitment among the employees. Their standard deviations range from.933 to1.414 which is not too far apart and really too bad. However the variances are from .870 to 2.00. According to the rating in this study the variances show that the variances are relatively big. In essence this means that employees have different levels of commitment to their jobs. A few outliers may be present which results in the relatively high variance of 2.00.
The values of turnover intentions are a bit higher in comparison with the rest. All the values are above 3.00 only one is above 3.5 the cut point that would allow us concludes that employees have high intentions to leave their current employer who happens to be the government.
55
Below is a table showing the frequencies of the demographics.
Table 4. 2. Frequency Results
56 Table 4.2 .(continued)
Total 133 100.0
Highest Educational Qualification
Graduate
Studies 33 24.8 28.4
28.4
College
Graduate 60 45.1 51.7 80.2
Diploma 23 17.3 19.8 100.0
Total 116 87.2 100.0
Missing
System 17 12.8
Total 133 100.0
57 4.3 Turnover Intentions (Model 1)
This sub-section discusses the empirical results for model 1 in which job satisfaction and employee commitment are tested against various items of turnover intentions as the dependent variable. Various tables are presented showing results of the analysis within model 1. Table 4.3 presents results for the first item of turnover intentions.
Table 4.3. Multiple Regression Analysis of job satisfaction and employee commitment and turnover intentions
I like my job better than the average worker
Y1= Turnover intentions. Q15 =I intend to remain in my current profession but to leave my current employer for a better one within these two years.
58
Turnover intentions based on job satisfaction and employee commitment for the next 2 years are shown in Table 4.3 above. By using backward elimination process, after examining the p–value for the 2 independent variables, the highest insignificant one is eliminated. The empirical results show that turnover as a function of Job Satisfaction variable factors and Employee Commitment variable factors. This process is repeated three times until in the third equation the remaining independent variables reach at least the target of 10% level of significance or even lower.
The parameter for each variable represents the percent change in the Turnover Intentions due to a 1% change in the independent variable (job satisfaction). Looking at Eq3, the regression reveals that 2 of the variables became significant at least at 1% level. For instance, the parameters of Job Satisfaction factors: only 1 of 2 variables is significant at 1%
level. The value of the parameter of liking the job (q11) is positive (0.607) with a t-ratio of 7.584. There is a statistical significance between liking one’s profession and leaving the current employer for a better one. There is a relationship between employees’ liking of the job and turnover intentions. We can therefore predict turnover intentions using this factor of job satisfaction. This means that employees like their jobs more and are willing to retain their careers or professions but have strong intentions to move to find another employer.
Thus there is satisfaction with the job itself but not the employer, hence the intent to quit the job and work elsewhere. In other words, the more employees like their professions, the more they want to work with good or better employers.
On the other hand, the parameter for Employee Commitment factors only 1 of 2 variables is significant at 1% level. The value of the parameter for employee commitment to their current employer with no intention to leave but can take advantage of a risen opportunity (q16) is positive (0.322) with a t-ratio of (4.027). This means that there is significant relationship between availability of new jobs and turnover intentions. The more job openings become available, the more employees will have intentions to move to the new jobs. We can therefore predict turnover intentions using this factor of commitment. Thus employees do not necessarily have the intentions to leave their current employer but may do so if an opportunity opens up. That is employees will remain committed to their current employer on the condition that no job opportunity opens up. Thus the employees’
commitment is short-lived since most of the people will desire to change employer in the
59
next 2 years. The rest of the parameters are not significant and will therefore not be discussed in great detail.
Finally the R-square stands at 0.803 after the elimination process; therefore the model can explain 80% of the variance in turnover intentions from job satisfaction and employee commitment.
Results for hypothesis
For the Job Satisfaction variables, the parameter for Q5, (the overall joy from the job as a whole) is not significant. Retain the null hypothesis: βYX = 0
There is no relationship between overall job satisfaction and turnover intentions.
Therefore turnover intentions cannot be predicted using this factor (the overall joy one gets from the job).
Results for hypothesis
For the Job Satisfaction variables, the parameter for Q11, (liking the job better than the average worker) is positive (0.607) with a t-ratio of (7.584) significant at level 1%. This means that there is a relationship between employees’ liking of the job and turnover intentions. We can therefore predict turnover intentions using employees’ liking of the job.
The more the employees like their job but the less employees like the employer, the higher the turnover intentions. Thus turnover depends largely on the liking of the employer and not just the job. That is whether the employees like the employer as a whole or not in terms of policies, HR strategies and others. Reject the null hypothesis, and accept the alternative:
βYX ≠ 0
There is a relationship between employee job satisfaction and dissatisfactions with employer, and turnover intentions.
60 Results for hypothesis
For the Employee Commitment variables, the parameter for Q8, (the employer offers better salary and other privileges, therefore cannot leave the job) is not significant at any level (1%, 5% and 10%). Retain the null hypothesis: βYX = 0
There is no relationship between job commitment and turnover intentions. Therefore, we cannot predict turnover using this factor variable (salary and privileges offered by the current employer).
Results for hypothesis
For the Employee Commitment variables, the parameter for Q16, (no intention to leave employer but do so if an opportunity arises) is positive (0.322) with a t-ratio of (4.027) significant at any level 1%. This means that there is a relationship between the availability of opportunities and turnover intentions. We can therefore predict turnover intentions using commitment to employer and availability of job opportunities. Therefore all conditions being equal the employee will remain at the job; but the more the opportunities for other jobs arise the more likely the employee will have intentions to leave the current employer.
Thus the more job opportunities arise; the more commitment to the job goes down and hence, more turnover intentions. Reject the null hypothesis, and accept the alternative:
βYX ≠ 0
61 4.4 Turnover Intentions (Model 1)
This sub-section presents the results of the regression analysis in which the independent variables (job satisfaction and employee commitment) are tested against the second item for the (dependent variable) turnover intention. A thorough discussion of the results follows at the bottom.
Table 4.4. Multiple Regression Analysis of job satisfaction and employee commitment on turnover
I like my job better than the average worker
Y1= turnover intentions. Q17 = (I am considering taking another job the soonest I find it)
62
A model of turnover intentions is shown in Table 4.4 by using backward elimination process, after examining the –value for the 2 independent variables, the highest insignificant one is eliminated. The table shows empirical results for turnover as a function of Job Satisfaction variable factors and Employee Commitment variable factors. This process is repeated three times until in the third equation, the remaining independent variables reach at least the target of 10% level of significance. It was found however that not all variables reached the 10% level. Only two factor variables reached a 1% level of significance with none reaching any other significance level.
The parameter for each variable represents the percent change in the Turnover Intentions due to a 1% change in the independent variable. Looking at Eq3, the regression reveals that 2 of the variables became significant at least at 1% level. For instance, the parameters of Job Satisfaction factors: only 1 of 2 variables is significant at 1% level. The value of the parameter of liking the job (q11) is positive (0.658) with a t-ratio of (7.360).
This means that there is a relationship between employees’ liking of the job and turnover.
We can predict turnover using this factor. It may mean that most of the employees like their jobs better than other employees however they are still considering taking another job the soonest they can find it. Thus liking the job does not necessarily mean maintaining the same job (employer).
The parameters of Employee Commitment factors: only 1 of 2 variables is significant at 1% level. The value of the parameter of commitment to the current employer with no intentions to leave although they may do if an opportunity opens up (q16), is positive (0.239) with a t-ratio of (2.671). This means that there is a relationship between employee commitment and turnover. We can predict turnover intentions using this factor for employee commitment. In this case it shows that employees have commitment to their jobs but are not fully committed to their current employer and do not necessarily have intentions of looking for another employer in the immediate future. However, they may take on a new job if an opportunity opens yet they wish to get a new job in the next 2 years. The broader picture here is that there is lack of serious commitment to the employer and hence the
63
intention to leave. The rest of the parameters are not significant and will therefore not be discussed in great detail.
Finally the R-square stands at 0.757 after the elimination process; therefore can explain 76% of the variance in turnover intentions from job satisfaction and employee commitment approach.
Results for hypothesis
For the Job Satisfaction variables, the parameter for Q5, (the overall joy from the job as a whole) is not significant at any level (1%, 5% and 10%). Retain the null hypothesis:
βYX = 0
There is no relationship between overall joys from the job and turnover intentions. We cannot predict turnover intentions using the overall joy that employees get from their jobs.
Results for hypothesis
For the Job Satisfaction variables, the parameter for Q11, (liking the job better than the average worker) is positive (0.658) with a t-ratio of (7.360) significant at level 1%. This means that there is a relationship between employees’ liking of the job and turnover. We can predict turnover using this factor. In this case it shows that the higher the employee likes the job better than the average worker, the higher the job turnover. This is contrary previous research findings that the higher the employees’ job liking, the lower the job turnover.
Reject the null hypothesis, and accept the alternative βYX ≠ 0
There is a relationship between employees liking of the job and turnover and we can predict turnover intention using employees’ liking of the job.
Results for hypothesis
For the Employee Commitment variables, the parameter for Q8, (the employer offers better salary and other privileges and so cannot leave job) is not significant at any level (1%, 5% and 10%) and is retained. This means that there is no relationship between the
64
employer’s offering of better salary and turnover intentions. Therefore we cannot predict turnover intentions using this factor. Retain the null hypothesis: βYX = 0
Results for hypothesis
For the Employee Commitment variables, the parameter for Q16, (no intention to leave employer but do so if an opportunity arises) is positive (0.239) with a t-ratio of (2.671) significant at any level but at 1% only. This means that there is a relationship between employee commitment and turnover intentions. We can therefore predict turnover intentions using employee commitment. All conditions being equal the employee will remain at the job;
but the more the opportunities for other jobs arise, the more likely the employee will have intentions to leave the current employer. Thus the more job opportunities arise; the more commitment to the job goes down and hence, more turnover intentions. Reject the null hypothesis, and accept the alternative: βYX ≠ 0
65 4.5 Turnover Intentions (Model 1)
This sub-section introduces the results of the regression analysis in which job satisfaction and employee commitment are the independent variables whereas the turnover intention is the dependent variable. A thorough discussion of the results follows at the bottom.
Table 4.5. Multiple Regression Analysis of job satisfaction and employee commitment on turnover intentions
I like my job better than the average worker
Y1 = Turnover intentions). Q18 =I have no intention to leave my current employer at all.
66
A model of turnover intentions is shown in Table 4.5. By using backward elimination process, after examining the p–value for the 2 independent variables, the highest insignificant one is eliminated. The table shows empirical results for turnover as a function of Job Satisfaction variable factors and Employee Commitment variable factors. This process is repeated three times until in the third equation the remaining independent variables reach at least the target of 10% level of significance. It was found however that not all variables reached the 10% level. Only two factor variables reached a 1% level of significance with none reaching any other significance level.
The parameter for each variable represents the percent change in the Turnover Intentions due to a 1% change in the independent variable. Looking at Eq3, the regression reveals that 2 of the variables became significant at least at 1% level. For instance, the parameters for Job Satisfaction factors: only 1 of 2 variables is significant at 1% level. The value of the parameter of liking the job (q11) is positive (0.474) with a t-ratio of (7.136).
This means that there is a relationship between how the employees like their job and turnover intentions. We can predict the employees’ intention to leave their current employer by using the factor of how employees like their jobs. The broader picture is that although employees like their jobs better than the average workers, there is still an indication that they still have intentions to leave their current employer.
The parameters for Employee Commitment factors: only 1 of 2 variables is significant at 1% level. The value of the parameter of commitment to the employer (q16) is positive (0.494) with a t-ratio of (7.452). This means that there is a relationship between commitment and turnover within if there is an opportunity. The explanation may be that the employees are not fully committed to their jobs hence do not necessarily think of leaving the
The parameters for Employee Commitment factors: only 1 of 2 variables is significant at 1% level. The value of the parameter of commitment to the employer (q16) is positive (0.494) with a t-ratio of (7.452). This means that there is a relationship between commitment and turnover within if there is an opportunity. The explanation may be that the employees are not fully committed to their jobs hence do not necessarily think of leaving the