This chapter focuses on critically identify the research hypotheses. The chapter is mainly divided into three parts, which are descriptive statistic analysis and inferential statistic analysis. The first section present is descriptive analysis of personal information of observers and exploring their perception towards politics issues in organization and supervisors as well as examines their satisfaction level on their careers. The second part shows the validity of the study. The last part emphasizes on the finding of research hypotheses by using hierarchical regression analysis.
Descriptive Statistics
The data was gathered from 300 participants who are full-time employees working as the first-lined customer representatives, which are check-in and boarding gate agents, at Suvarnabhumi airport in Bangkok, Thailand. To collect the information of the participants, gender, age, job position, tenure, department, English proficiency, and supervisor’s nationality were included. Female is the majority participants in the study (64.7%). Most of them ages between 21 – 30 years (66.3%). There are averagely distributed in job position, which consist of agent, leading agent, and supervisor. Regarding with frequency, most of them are agent (37%) whose responsibility are at both check-in counter and boarding gate (49.3%). Their service length in present companies is principally between 1 – 3 years (39%).
Most of them use English proficiency test of TOEIC (82%). In terms of supervisor, most of them are Thai (60%). The frequency and percentage of the demographic information are summarized in the Table 4.1 – 4.2.
Table 4.1.
Descriptive Statistics (n = 300)
Item Sample characteristics Frequency Percentage
Gender Male 106 35.3
Table 4.2.
Supervisor’s Nationality (n = 300)
Item Sample characteristics Frequency Percentage
Supervisor's nationality Thai 180 60.0
Korean 21 7.0
According to Zikmund (2003), reliability refers to the overall consistency of measurement and stability of overall finding. Cronbach’s alpha coefficient was selected in order to guarantee the stability and consistency of questions of each variable. The greater value of alpha, it explains the high consistency. Nunnally (1978) suggested the rule of thumb for Cronbach’s alpha value should be greater than 0.70 for indicate term of reliability. After the
and perceived supervisor support was 0.94. The result is shown in the Table 4.3.
Table 4.3.
Reliability Analysis
Variable Cronbach's Alpha N of Items
Organizational politics .54 12
Job satisfaction .91 20
Perceived supervisor support .94 16
Confirmatory Factor Analysis
The researcher selected the AMOS for performing a validity test of the questionnaire that will be used for collecting data. By doing so, the researcher was able to confirm if the questionnaire was suitable with the measurement method and fit the previous theories of this study. In this study, the researcher adopted various indicators to measure the validity of organizational politic, job satisfaction, and perceived supervisor support as follow:
Firstly, Chi-square (χ2) test was utilized to measure the fitness of overall model between expected and observed variables covariance patterns. In the case of the chi-square degree of freedom ratio, smaller rather than larger values indicate a good fit (Jöreskog and Sörbom, 1989). This is also suggested it should have ratio lower than 3 for indicating the good model fit.
Secondly, Root Mean Square error of approximation (RMSEA) was additionally implemented for supporting the fit in the large sample size. The value of RMSEA near to zero is explaining the perfect model fit. Steiger (1990) stated the values of RMSEA above 0.10 shows an inadequate fit, while values below 0.05 indicate a very good fit.
Thirdly, Normed Fit Index (NFI) was tested for showing the difference in chi-square value of hypothesis model and chi-square from the null model (Bentler & Bonett, 1980). But NFI is having disadvantage by no penalty for adding parameter in the model, which explains
the negative bias. The non-normed fit index (NNFI) was introduced to solve the problem of bias, which is later called Tucker-Lewis index or TLI. Another incremental measurement is called comparative fit index (CFI) for finding the discrepancy between value of hypothesis model and the null model. Both of NFI and NNFI have ranged between 1 and 0, where the cut point should be higher than 0.90 for examining the good index (Bentler, 1990).
According to the Table 4.4, there are 12 items of organization politic, 20 items of job satisfaction and 16 items of perceived supervisor support. The results of fit indexes were explained as followed: Chi-square (χ2) of organization politic was 82.12, it presented good fit indexes, where the rest of the indexes were: NFI = .92, CFI = .93, TLI = .81 and RMSEA = .18. Chi-square (χ2) of job satisfaction was 249.77, it presented good fit indexes, where the rest of the indexes were: NFI = .91, CFI = .93, TLI = .87 and RMSEA = .10. Chi-square (χ2) of perceived supervisor support was 315.60, it presented good fit indexes, where the rest of the indexes were: NFI = .91, CFI = .93, TLI = .91 and RMSEA = .09.
Table 4.4.
Result of Confirmatory Factor Analysis (n = 300)
χ2 df χ2/df NFI CFI TLI RMSEA
Organizational Politics 82.12 21 3.91 .92 .93 .81 .18
Job Satisfaction 249.77 120 2.08 .91 .93 .87 .10
Perceived Supervisor Support 315.60 120 2.63 .91 .93 .91 .09
Pearson Correlation Analysis
In term of measuring the relationship among organizational politics, job satisfaction, and perceived supervisor support, the researcher selected Pearson correlation analysis in order to examine the strength and sign of relationship among all variables. The result is shown in the Table 4.5. Organization politics was negatively correlated to job satisfaction (r
= -0.60, p < 0.01) and perceived supervisor’s support also had negative relationship with
organization politics (r = -0.41, p < 0.01). Furthermore, the perceived supervisor support was high correlated with job satisfaction (r = 0.74, p < 0.01).
Table 4.5
Mean, Standard Deviation and Correlation (n = 300)
Mean S.D 1 2 3 4 5 6
1. Job Position 1.95 .83
2. Job Tenure 2.25 .82 .63**
3. English Proficiency – .84 -.06 .07
4. Organizational Politics 3.32 1.20 -.06 .09 .02 (.54)
5. Perceived Supervisor Support 3.44 .82 -.04 -.06 -.05 -.41** (.91)
6. Job Satisfaction 3.36 .71 .06 -.07 -.07 -.60** .74** (.94) Note: *p<.05, **p<0.01
Hierarchical Regression Analysis Organizational Politics and Job Satisfaction
To test out the effect of organizational politics on job satisfaction, the result of hierarchical regression analysis in the Table 4.6 was applied for testing the hypothesis 1. In the initial step, the control variables (job position, job tenure, English proficiency) were entered. In the step 2, the independent variable (organization politics) was entered in order to test the hypothesis.
Hypothesis 1 examined that organizational politics has a significantly negative effect on job satisfaction. In the Model 1, all of control variables showed no effect on job satisfaction (p >
0.05). In the Model 2, organizational politics were added to measure the association with the job satisfaction. The result found organizational politics was significantly having negative effect on job satisfaction (β = -0.30, p < 0.01). It implied that full-time employees who perceived high level of organization politics; they will have low satisfaction level on their careers. Thus, Hypothesis 1 was supported.
Table 4.6.
Result of Regression Analysis for Job Satisfaction (n = 300)
Variable Job satisfaction
The Moderating Effect of Perceived Supervisor Support
To test out the moderate effect of perceived supervisor support on the relationship between organizational politics and job satisfaction, the result of hierarchical regression analysis in the Table 4.7 was applied for testing the Hypothesis 2. In order to test the hypothesis, there are three steps in the regression analysis. In the initial step, the control variables (job position, job tenure) were entered. In the step 2, the independent variable (organization politics) and moderator (perceived supervisor support) were entered. The last step is explaining the interaction term of organization politics and perceived supervisor support was entered. Hypothesis 2 examined that perceived supervisor support helps weakening the negative relationship between organizational politics and job satisfaction. In the Model 1, all of control variables showed no effect on job satisfaction (p > 0.05). In the
Model 2, organization politics and perceived supervisor support were added to measure the association with the job satisfaction. The result found organizational politics was significantly having negative effect on job satisfaction (β = -0.120, p < 0.05), while perceived supervisor support was a significantly positive relationship with job satisfaction (β = 0.496, p
< 0.05). It can be implied that full-time employees who perceived supervisor support help weakening the negative relationship between organizational politics and job satisfaction.
However, the Model 3 explained the interaction of organization politics and perceived supervisor support has no significant effect on job satisfaction, but it has a negative relationship (β = -0.123, p > 0.05). Thus, Hypothesis 2 was not supported.
Table 4.7.
Result of Regression Analysis for Perceived Supervisor Support as Moderator (n = 300)
Variable Job satisfaction
ΔF 0.53 218.99** 0.98 Note: *p<.05, **p<0.01
Table 4.8.
Result of Hypotheses Testing
Hypotheses Result
H1 Organizational politics has a negative effect on job satisfaction Supported H2 Perceived supervisor support helps weaken the negative
relationship between organizational politics and job satisfaction
Not Supported Table 4.7. (continued)