In this chapter, the correlation analysis was used to gain a preliminary knowledge of the relationship between variables. In addition, hierarchical regression and K-mean cluster analysis were conducted to test the hypotheses proposed in this study.
Correlation Analysis
Mean, Standard deviation, Cronbach’s alpha and the correlation between variables are presented in Table 4.1. Starting from the control variables, as literature stated, Age was significantly and negatively correlated with turnover intention (r=-.33, p<.01), meaning that the older an employee is, the less likely they will think about quitting. Marital status and the number of dependents were both significantly and negatively correlated with turnover intention as well (r=-.26, p<.01; r=-.23, p<.01). It means that when people are married or when they have more dependents relying on them, their intention to leave their job will be lower. When employees are managers in their companies, they tend to stay at the position rather than leaving (r=-.13, p<.05). In addition, employees’ education level was significantly and positively correlated with turnover intention. This indicated that people with higher degrees are more likely to think about leaving their company.
Pearson’s Correlation was also used to examine the relationship between independent variables, dependent variables and moderators. In this research, the data did show that career plateau, and its two dimensions: job content plateau and hierarchical plateau were significantly and positively correlated with turnover intention (r=.37, p<.01; r=.16, p<.01; r=.35, p<.01). These suggested that when employees meet a stagnant situation no matter on their job content or their position
in companies, they tend to have a higher intention to leave the organization that makes them stuck at the current situation. However, the results of Pearson’s correlation were inconclusive and did not test the hypothesized effects of career plateau on turnover intention. Therefore, hierarchical regression was utilized in the following sections to test the proposed research hypotheses.
47
Table 4.1.
Mean, Standard Deviation, Correlation and Reliability among Research Variables (n=412)
Mean Std. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Note: **. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).
Gender: 0= Male; 1= Female 0= Non-Managerial Positions; Managerial Position: 1= Managerial Positions Marital status: 0=Non-Married; 1=Married Education Level: 1=High school or under; 2=Vocational School; 3=Bachelor degree; 4=Master degree; 5=PhD degree
Numbers in the parentheses represent the values of Cronbach’s Alpha.
Cluster Analysis
In this research, cluster analysis was conducted to investigate the distinct segments of the participants. K-mean cluster algorithm was utilized to obtain three final clusters, which were later on named as High-Career Pursuer, Mid-Career Pursuer and Low-Career Pursuer. The final cluster centers are shown in Table 4.2. and Figure 4.1. demonstrates the line graph of the three clusters.
Table 4.2.
Final Cluster Centers
Cluster
Low-Career Pursuer High-Career Pursuer Mid-Career Pursuer
Autonomy / Independence 2.47 5.09 4.45
Stability / Security 3.23 4.99 4.53
Technical / Functional Competence 2.46 5.27 4.59
General Management Competence 2.19 4.18 2.94
Entrepreneurial Creativity 1.71 5.03 3.23
Service / Dedication 2.11 5.21 4.29
49
(1) The Low-Career Pursuer: Participants in this cluster scored relatively lower than the other two clusters on all eight anchors. Their average age was 33.53 years old and their career year was 8.42, which were both the lowest among the clusters.
Among them, 57.9% were entry-level employees and 26.3% were at the first-level managerial positions, meaning that these participants were either not interested in pursuing a higher career or they were still at the very start of their career. The latter case suggested that they were still exploring in their career path and do not know what were the essentials they want to pursue. From the pattern, it can be noted that employees in this cluster valued more on the stability and job security, as well as the work-life balance in their career.
(2) The Mid-Career Pursuer: In this cluster, the participants have their scores of 8 anchors between the Low-Career and High-Career cluster. The pattern looked similar to those in High-Career cluster; however, they scored relatively low in the general management competence and entrepreneurial creativity. Their average age was 39.51 and the career year was 15.26. A 52.9% of these employees were in the non-managerial positions; 18.6% were first level managers; 15.2% were mid-level managers and 6.4% are in high managerial positions. These participants had a better understanding about what they wanted for their career and knew quite well that they did not want to or were not ready to be general managers or create their own business.
(3) The High-Career Pursuer: As to people classified as High-Career Pursuers, they tended to pursue as many anchors as they can. However, the result still shows that people would rather be technical or functional professionals in specific domain than being at general managerial positions. Their average age was 42.55 and the average career year was 18.16. Among all participants in this cluster, only 37.6% were not in managerial positions; 16.4% were first level managers; 21.7% were in mid-level managerial positions and 12.7% were high level managers. The result suggested that most of these people had a relatively mature career and they wanted to attend to each and every aspect of their career.
The cluster profiles are shown in Table 4.3.
Table 4.3.
Average Number of Dependent 1.37 1.31 1.72
Marital Status
51
Hierarchical Regression Analysis
In this section, hierarchical regression was performed to test the research hypotheses. For career plateau, the entire variable and its both dimensions were analyzed separately. The results are shown in Table 4.2.
In Table 4.4., Model 1 showed that control variables were analyzed and only age had a significant negative effect on turnover intention (β=-.235, p<.001). Model 2 tested the first research hypothesis: Career plateau is positively related to employees’
turnover intention. This hypothesis was supported by the result that career plateau is significantly positively influencing turnover intention (β=.374, p<.001). In Model 3 and 4, the researcher moved on to the dimensional level of the independent variable and found that both job content plateau and hierarchical plateau had a significant positive effect on turnover intention (β=.183, p<.001; β=.335, p<.001). The results supported the Hypothesis 1 that career plateau had a significantly positive effect on turnover intention.
In Table 4.5., the sample was divided in to three groups according to the three clusters generated in the previous section. The researcher intended to investigate the differential effects of participants from the three clusters regarding how their career plateau affected their intention to turnover. For Low Career Pursuers, there was no significant effect of career plateau on turnover intention (β=.083, p>.05). However, for Mid-Career Pursuers and High Career Pursuers, the results showed that their career plateau did significantly positively influenced their turnover intention (β=.358, p<.001; β=.400, p<.001) and that the effect for High Career Pursuers was stronger than that of Mid-Career Pursuers (.400>.358).
Table 4.5.
Hierarchical Regression Result among Clusters (n=412)
Variable Turnover Intention
Low Career Pursuers Mid-Career Pursuers High Career Pursuers
β β β
53
Later on, the researcher went on analyzing the dimensional level for the reason that the two dimensions: job content plateau and hierarchical plateau have distinct differences. The participants might experience them quite differently. Therefore, the dimensions were tested to see whether there are indeed variations when it comes to the effects of turnover intention. In Table 4.6., turnover intention was first regressed on job content plateau. It demonstrated that for Low Career Pursuers, job content plateau did not affect their turnover intention (β=.029, p>.05). For Mid-Career Pursuers, job content plateau did have a significant positive influence on turnover intention (β=.250, p<.001). As to High Career Pursuers, job content plateau also had a significant positive effect on turnover intention (β=.166, p<.05); however, the effect was weaker and less significant compare with that of Mid-Career Pursuers.
Table 4.6.
Hierarchical Regression Result among Clusters (n=412): Turnover Intention Regressed on Job Content Plateau
Variable Turnover Intention
Low Career Pursuers Mid-Career Pursuers High Career Pursuers
β β β
Finally, the researcher analyzed the effects of hierarchical plateau on turnover intention from the three clusters. Same as the previous results, for Low Career Pursuers, hierarchical plateau did not have any significant effect of turnover intention.
While the analysis on Mid-Career Pursuers and High Career Pursuers found that hierarchical plateau did have significant positive effect on turnover intention (β=.250, p<.001; β=.399, p<.001) and the effect seemed to have stronger influence for High Career Pursuers than Mid-Career Pursuers. The results are presented in Table 4.7.
Table 4.7.
Hierarchical Regression Result among Clusters (n=412): Turnover Intention Regressed on Hierarchical Plateau
Dependent Variable Turnover Intention
Low Career Pursuers Mid-Career Pursuers High Career Pursuers
β β β
Hierarchical Plateau .133 .250*** .399***
Adj. R2 .515 .478 .073 .129 .169 .322
ΔR2 .005 .059 .152
F for ΔR2 3.098* 3.358* 3.655** 5.283*** 7.351*** 13.733***
n 19 19 204 204 189 189
*p<.05. **p<.01. ***p<.001.
55
To sum up, based on the results of hierarchical regression analysis used to test the research hypotheses, Hypothesis 1 and 2 were both supported. That is, career plateau did have significant positive effect on turnover intention and career anchor profiles did moderate this relationship. The integrated result is shown in Table 4.8.
Table 4.8.
Hypotheses Testing Result Summary
Hypotheses Result
Hypothesis 1
Career plateau will be positively significantly related
employees’ turnover intention. Supported
Hypothesis 2
Career anchor profiles will moderate the relationship
between career plateau and turnover intention. Supported
56