This chapter presents the outcomes of various data analyses conducted for this study.
One-Way ANOVA
In order to answer one of the research questions of this study (“Are there significant differences among for-profit, nonprofit and public sector employees’ job satisfaction when looking at perceived supervisor support and customer feedback as independent variables?”), One-Way ANOVA was performed. The descriptive statistics for the comparison of employee job satisfaction among workers of different sectors are shown in Tables 4.1.
Table 4.2 indicates statistical significance of employee job satisfaction between groups, (F(2,310) = 9.72, p<.001). Due to the fact that the study split the work sector into three categories, Scheffe’s Post Hoc Test was needed to compare the means of employee job satisfaction among business, public and nonprofit respondents. Observation of Table 4.3 shows the mean difference of employee job satisfaction between business (M=3.41, SD=.71) and public sector (M=3.56, SD= .76) employees as insignificant. This demonstrates that there are no significant levels of differences in job satisfaction among employees in the business and public sectors who participated in this study. In addition, the mean difference between public (M=3.56, SD=.76) and nonprofit sector employees (M=3.89, SD=.59) also show no statistical significance. Contrastingly, the mean difference between business (M=3.41,
Table 4.1.
Descriptives of Employee Job Satisfaction among Work Sectors
Working sector N Mean Std. Deviation Std. Error
Business sector 221 3.41 .71 .047
Public sector 43 3.56 .76 .12
Nonprofit sector 49 3.89 .59 .08
Total 313 3.51 .72 .04
Results thus answered the fourth research question of this study; there is a significant difference in levels of job satisfaction between nonprofit and business sector employees, but no statistically significant difference between public and nonprofit or public and business sector employees. Consequently, nonprofit respondents of this study show to have a higher level of job satisfaction in contrast to respondents from the business sector. Therefore, this test concluded that nonprofit respondents seem to be more satisfied with their jobs than those working in the business sector.
Correlation Analysis
Through Pearson’s correlation analysis, the results revealed both expected and unexpected findings depicted in Table 4.4. As theorized, there is a strong positive correlation
SD=.71) and nonprofit sectors (M=3.89, SD=.59) is significant.
Table 4.3.
Scheffe’s Post Hoc Test
Dependent Variable (I) Work Sector (J) Work Sector
One Way Analysis of Variance in Employee Job Satisfaction among Work Sectors
Sum of Squares df Mean Square F Sig.
Between Groups 9.42 2 4.71 9.72 .000
Within Groups 150.15 310 .48
Total 159.56 312
between perceived supervisor support and employee job satisfaction (r=.70, p<.001).
Therefore the sample reflects the notion that the more an individual believes to have support from his or her supervisor, the higher job satisfaction they indicated as having. There is also a moderate correlation between perceived customer feedback and employee job satisfaction (r=.30, p<.001); meaning when an employee perceived to receive more positive customer feedback, the higher satisfaction with their job they reported to have. In addition, high correlation between work motivation and employee job satisfaction (r=.62, p<.001) suggests that the higher motivation an employee has, the more satisfied he or she feels at their current job position.
Perceived supervisor support and work motivation (r=.51, p<.001), and customer feedback and work motivation (r=.40 p<.001) also show significant positive correlations. In this study, employees who felt they received more support from their supervisors and/or received more feedback from their customers also reported to feel more motivated about their work.
Apart from Pearson’s correlation analysis yielding some expected outcomes, there were also some unexpected outcomes. A notable positive correlation includes perceived supervisor support and perceived customer feedback (r=.21, p<.001). Results suggest that those who believe to have more supervisor support also believe to receive more customer feedback.
Correspondingly, the relationship between familiarity with customers and job satisfaction (r=.27, p<.001) produced a significant correlation. It appears the more an employee knows about his or her customer, the more satisfied he or she feels about their job. From a different perspective, when employees are more satisfied, they are inclined to pay more attention and get to know their customers. Perhaps one of the more interesting correlations occurred between level of education and job satisfaction (r=.01, n.s.) and work motivation (r=.06, n.s.).
It appears higher or lower levels of education give no significant indication as to whether people are work motivated or satisfied with their current jobs. In other words, those with
higher education in this study did not indicate as having more job satisfaction and vice versa.
Continuing to examine level of education, results indicate a significant positive correlation between level of education and nonprofit employee workers (r=.25, p<.001).
Individuals reporting as working in the nonprofit sector also reported as having higher levels of education. Yet on the other hand, males correlated negatively with levels of education (r=-.20, p<.001) and negatively with the nonprofit sector (r=-.12, p<.05). These correlations imply male respondents as having lower levels of education and employment in the nonprofit sector.
Other notable negative correlations included relationships between customer service agents (CSA) and (1) employee job satisfaction (r=-.32, p<.001), (2) perceived supervisor support (r=-.33, p<.001), and (3) work motivation (r=-.12, p<.05). All in all, responses from customer service agents in majority revealed to be less satisfied with their jobs, perceived to have less support from their supervisors in comparison to others, and felt less motivated about work. Full details of Pearson correlations are depicted in Table. 4.4.
From Table 4.4, research question 2 (“Is employee job satisfaction more positively associated with perceived customer feedback or perceived supervisor support”) is answered.
Results show perceived supervisor support to be highly correlated with employee satisfaction (r= .70, p< .001), while perceived customer feedback is moderately correlated with employee job satisfaction (r=.30, p<.001). Although the table shows many significant positive and negative correlations between variables, hypotheses and remaining research questions could not be confirmed nor denied through correlation. More advanced statistical methods were required to test the direct hypotheses.
Structural Equation Modeling (SEM)
As was mentioned in chapter 3, all items from the study underwent CFA for respectable variables. The modified list of items for the entire study was then inputted into SPSS AMOS to produce a structural model. An illustration is presented in figure 4.1.
Figure 4.1. Structural equation model.
Path analysis revealed standardized beta coefficients, r-squares, and model fit summaries to test the study’s hypotheses. Although some SEM regression weights display loadings below the .5 cutoff, all item loadings were significant in both CFA and SEM outputs, therefore this model was used.
The paths and beta coefficients of most interest for this study are those between latent variables – perceived customer feedback to employee job satisfaction; perceived supervisor support to employee job satisfaction; work motivation to employee job satisfaction;
perceived supervisor support to work motivation; and perceived customer feedback to work motivation. These five paths exposed answers to hypotheses of this research. The control variables included in the SEM were selected based on correlations with the latent variables and previous literature reviewed. The selected control variables included gender and education completed.
From SEM outputs of X2/df and RMSEA in Table 4.5 indicate acceptable model fit, however GFI and AGFI are slightly lower than ideal in the default model’s model fit summary. In addition, RMR appears above the desirable cutoff. Nonetheless, measurement model figures showed to be adequate in progressing with the analysis. From the SEM, standardized regression weights (beta coefficients) were yield. As depicted in Table 4.6, the beta coefficients between perceived supervisor support and work motivation (ß=.49, p<.001), perceived customer feedback and work motivation (ß=.37, p<.001), and work motivation employee job satisfaction (ß=.30, p<.001) are all statistically significant. In addition, the
Table 4.5.
Model Fit Summary of Structural Model
Model X2 df P X2/df RMR GFI AGFI RMSEA
Default model 1415.50 735 .000 1.93 .11 .82 .80 .05
regression path between perceived supervisor support and employee job satisfaction (ß=.63, p<.001) also shows high statistical significance. From these regression weights, the outcome illustrates causal effects between the mentioned variables. Perceived supervisor support and perceived customer feedback appear to have a direct positive influence on work motivation.
In turn, work motivation and perceived supervisor support appear to directly, positively influence employee job satisfaction. However, the direct relationship between perceived customer feedback and employee job satisfaction shows no significance. Full illustration of the SEM’s regression pathways are depicted in Table 4.6.
While Table 4.6 depicts only the direct effects of the SEM’s regression weights, Table 4.7 gives a more detailed description of the beta pathways. It illustrates standardized direct, indirect and total effects of the SEM along with the r-squares of work motivation and employee job satisfaction. A summary of the effects is listed in Table 4.7.
Table 4.6.
Standardized Coefficients of Structural Equation Modeling
Path Standardized coefficient
PSS ! WM .49***
PCF ! WM .37***
PSS ! EJS .63***
PCF ! EJS .07
WM ! EJS .30***
Gender ! EJS .02
Education Completed ! EJS -.02
Note. *p<.05, **p<.01, ***p<.001
Summary of Analyses Results
From the statistical analyses that were run for this research, all research questions and hypotheses were answered. Pearson’s correlation analysis revealed employee job satisfaction to be more positively associated with perceived supervisor support versus perceived customer feedback (research question 2). One-way ANOVA shows statistically significant differences in employee job satisfaction between nonprofit and business sector employees. However, between business and public, and nonprofit and public sectors, analysis reveal no significant differences in employee job satisfaction; thus answering research question 4.
The remaining research questions and all hypotheses were answered through structural equation modeling. By looking at the beta coefficients from SEM, perceived supervisor support appears to have more influence on employee job satisfaction as opposed to perceived Table 4.7.
customer feedback (research question 1). Work motivation was also found to positively influence employee job satisfaction (research question 3).
After running SEM, analysis shows partial support for Hypothesis 1. There is a direct positive influence of perceived customer satisfaction on employee job satisfaction, however the direct effect is insignificant. In contrast, the regression of perceived supervisor support on employee job satisfaction confirms high positive influence on employee job satisfaction at a statistically significant level. As for hypotheses 3 and 4, Table 4.6 illustrates significant moderate influence of perceived customer feedback and perceived supervisor support on work motivation. In turn, work motivation’s regression weight on employee job satisfaction is also moderate at a significant level; thus supporting hypothesis 5. A summary of hypotheses results is shown in Table 4.8.
Table 4.8.
Hypotheses Testing Results Summary
Hypotheses Result
H1 Perceived customer feedback positively influences employee job satisfaction.
Partially Supported H2 Perceived supervisor support positively influences
employee job satisfaction. Supported
H3 Perceived customer feedback positively influences work
motivation. Supported
H4 Perceived supervisor support positively influences work
motivation. Supported
H5 Work motivation positively influences employee job
satisfaction. Supported