This section covers the methodology of this research. It contains the research framework, research hypotheses, the sampling, the instrumentation, data collection, and data analysis. This research adopted the quantitative research process to investigate the relationships among psychological contract breach, career plateau and intention to leave.
Research Framework
The independent variable is psychological contract breach. The intention to leave is the dependent variable. The career plateau serves as the moderator to examine the relationship between the psychological contract breach and the intention to leave. The career plateau consists of two dimensions which are hierarchical career plateau and job content career plateau. Figure 3.1 shows the research framework.
Hypotheses
Based on the previous literatures, the research purpose, the research questions, and the hypotheses are described as follow:
Hypothesis 1. Psychological contract breach is positively associated with intention to leave.
Hypothesis 2. Career plateau positively moderates the relationship between psychological contract breach and intention to leave.
Higher career plateau strengthens the relationship between psychological contract breach and intention to leave.
Hypothesis 2a. Hierarchical career plateau positively moderates the relationship between psychological contract breach and intention to leave.
Higher hierarchical career plateau strengthens the relationship between psychological contract breach and intention to leave.
Hypothesis 2b. Job content career plateau positively moderates the relationship between psychological contract breach and intention to leave.
Higher job content career plateau strengthens the relationship between psychological contract breach and intention to leave.
Research Design Sample
The sample of this research was the employees who work at the bank in Taiwan. There are two main types of banks in Taiwan which are domestic banks and foreign banks. There were 38 domestic banks and 30 local branches of foreign banks in Taiwan in 2012 (Banking Bureau of Financial Supervisory Commission, R.O.C)
The term, Domestic Bank, as used in this Banking Act, shall mean an organization formed and registered in accordance with the provisions of Banking Act for purposes of transacting a banking business (The Banking Bureau of Financial Supervisory Commission, R.O.C, 2012). It included commercial Bank and Specialized Bank. On the other hand, the term, Foreign Bank, as used in this Banking Act, shall mean a Bank organized and incorporated in accordance with the laws of a foreign country, which Bank has been recognized by the government of the Republic of China and registered for business as a branch office within the territory of the Republic of China, in accordance with the Company Law and this Banking Act (The Banking Bureau of Financial Supervisory Commission, R.O.C, 2012).
In the past decade, plenty of graduated students were willing to work in the banks. Based on the numbers of banking employees from Directorate-General of Budget, Accountings and Statistics, Executive Yuan, R.O.C., there were around 127,000 banking employees in 2002 and there were only around 140,562 banking employees in 2012. Besides, there has been a high turnover in the banking every year.
Through the well-known human resource agencies like 104, 1111 human resource agency and the official website of each bank, it is easily to find there are plenty of openings in the banking. That is the reason why the banking employees are the sample in this study.
Commission, R.O.C., 90 % of head offices are located in the north and central of Taiwan. Therefore, the questionnaires were distributed to the domestic banks located in north and central of Taiwan instead of governmental banks such as Bank of Taiwan, Land Bank of Taiwan. Since the employees who work in the governmental banks have higher benefits and salaries than other domestic banks. Also their benefits are just similar as other civil servants in Taiwan. So it is hard for them to have intention to leave the governmental banks.
Data Collection
The sample of this study was 444 employees worked in 15 domestic banks in Taiwan (Appendix B). The convenient sampling was used in this research. The total distribution number of questionnaires was 790. In two months, 529 questionnaires were returned, 85 were invalid. The response rate of questionnaire was 56.2%.
Two main methods were used to conduct data collection. In the first method, responses from this study sample were obtained through a visiting to each bank branch in person and contacted with bank manager, who was in charge of collecting data from their employees. Then three steps were taken to conduct the data collection:
First, obtaining permission from the bank manager of each branch. Second, establishing the data collection schedule and deadline. Last, distributing the paper-based questionnaire.
The second method was a snow-ball sampling strategy. This method contained five steps to collect the data: first, identifying the potential subjects who work in the banks. Second, contacting and sending the request to the potential subjects. Third, subjects nominating another person with same characteristics. Fourth, establishing the data collection schedule and deadline. Last, distributing the paper-based questionnaires.
Measurement Psychological Contract Breach
Morrison and Robinson (1997) defined that perceived psychological contract breach referred to an employee discovering there was unfairness in the social exchange process, and an employee did not receive expected results for accomplishing his or her obligations from an organization. This research adopted the psychological contract breach scale which was based on Robinson and Morrison (2000). The original Cronbach’s alpha reliability of the psychological contract breach in this study was 0.86. The scale contained a total of 5 items. An example item is “I have not received everything promised to me in exchange for my contribution.” The questionnaire asked participants to self-report the real situation while they worked in the organization. Responses were designed on a 7-Point Likert scale, Respondents were asked to indicate the extent to which they agreed or disagreed with the following statements ranging from 1(strongly disagree) to 7 (strongly agree). The high score of scale indicate that the person perceives higher psychological contract breach.
Career Plateau
Career plateau was defined as the point in one’s career at which the probability of a vertical assignment or a promotion in the near future is unlikely. (Ference, Stoner,
& Warren, 1977) This research adopted the career plateau scale which was based on Allen et al., (1999). The scale contained in total 12 items which are separated into two dimensions: hierarchical career plateau and job content career plateau. The original Cronbach’s alpha for this scale was 0.83 on hierarchical career plateau and 0.85 on job content career plateau. There are six items on hierarchical career plateau. An example item is “The opportunities for upward movement are limited.” And the other six items belong to job content career plateau, the items include like “My job tasks
self-report the real situation while they work in the organization. Responses are designed on a 5-Point Likert scale. Respondents were asked to indicate the extent to which they agreed or disagreed with the following statements ranging from 1(strongly disagree) to 5 (strongly agree). The high score of scale indicates that the person perceives the career plateau in the organization.
Intention to Leave
Intention to leave was defined as an individual thinking of quitting which is the next logical step after experienced dissatisfaction in the organization and intention to leave may be the last step prior to actual quitting (Porter & Steers, 1973). This research adopted the intention to leave scale which was developed by Meyer et al., (1993). The original Cronbach’s alpha for this scale was 0.83. The scale included three items. Items are “I frequently think about leaving my current employer.”, “I am likely to search for a job in another organization.” And “I am likely to leave the organization within the next year.” The questionnaire asked participants to self-report the real situation while they work in the organization. Responses are designed on a 7-Point Likert scale. Respondents were asked to indicate the extent to which they agreed or disagreed with the following statements ranging from 1(strongly disagree) to 7 (strongly agree). The high score of scale indicates that the person has higher intention to leave the organization.
Control Variables
According to the previous literatures, researchers found that some demographic variables had the impact on intention to leave. Therefore, these control variables would be included in the final section of the questionnaire to clarify the relationship among dependent, moderator variable, and independent.
Age
Mobley, Horner and Hollingsworth (1978), Arnold and Feldman (1982), Cotton and Tuttle (1986) found that the elder employees tended to have lower intention to leave than younger employees. The other studies related to the intention to leave issues believed that age was a significant effect, and it was found that age would have a negative relationship with intention to leave (Sager, Futrell & Varadarajan, 1989).
Therefore, the age was the control variable in this study.
Education
In the previous researches, education was viewed as an important variable that has impact on intention to leave researches. Mobley, Horner and Hollingsworth (1978) found that education had negative relationship with intention to leave. However, Cotton and Tuttle (1986) found that an employee with higher education had higher intention to leave the job than those with lower education. Therefore, the education was the control variable in this study.
Job Tenure
Arnold and Feldman (1982) found that job tenure had negative relationship with intention to leave in the previous researches. The other studies related to the intention to leave issues believed that job tenure was a significant effect (Abelson, 1986; Cotton
& Tuttle, 1986; John & Jeffery, 1986; Mobley et. al., 1978). Therefore, the job tenure was the control variable in this study.
Questionnaire Design
The scales adopted in this questionnaire were all mature scales. The content of scale’s items was all valid. The original scales are English version. To consider that most participants are Taiwanese, it would be more efficient for them to answer the questionnaire in Chinese. Therefore, the questionnaire in this research was translated into Chinese version. The researcher invited a high classmate to translate the original English version into the Chinese version. Then, three graduate students were invited to translate Chinese version into English version in order to make sure the Chinese version corresponded with the original English version.
All variables’ scale were put into one questionnaire, so the independent and dependent variables were answered by the same person and at the same period of time.
It may cause the problem of common method variance (CMV). In order to avoid the problem of CMV, the researcher made the effort to minimize as following:.
First, based on the suggestion from Podsakoff, MacKenzie, Lee and Podsakoff (2003), the study purpose and variables were not shown and clearly explained in the questionnaire. Second, the different point of Likert scales were used in the questionnaire. For example, psychological contract breach and intention to leave adopted 7-point Likert scale, and career plateau adopted 5-point Likert scale. Third, this study adopted reverse items in the questionnaire to prevent participants from filling in the questionnaire randomly.
Back Translation
The original scales are English version. To consider that participants are all Taiwanese, the researcher translated the scales into Chinese. In order to know whether the Chinese translation conformed original meaning of English, we conducted the back translation.
The researcher found friend Jams Lei who has studied abroad in the United
States since 2007. He majored in Business Administration at University of Oregon in the U.S. As for his English ability, he got a score of 97 in TOEFL. He translated the English version scales back to Chinese. Based on his translation, three graduate students who majored English before and one of them obtained 110 TOEFL and 980 TOEIC score, the others also obtained 800 TOEIC score translated Chinese into English. The main purpose was to ensure Chinese version and English version were in the same meaning. The researcher also asked the thesis advisor to confirm both versions to make sure the words translated into the consistent meanings.
Pilot Test
This research was a quantitative research. The researcher did the pilot test before collecting the data. In order to understand the samples, to check the reliability of scales and examine the relationship between all variables, pilot test was distributed to 50 banking employees who work in E. SUN bank. Only 43 questionnaires were valid and 7 questionnaires were invalid.
First, from the demographic information about the samples, half of the 43 participants were female (58.1%). Most of them were single (74.4%) and aged from 26 to 30 years old (41.9%). When it came to the education background, over 72.1%
participants had bachelor degree. Regarding their working position, most of them were first-line employees (88.4%). The participants’ total tenure in banking industry was principally for 1 to 5 years (51.2%). The frequency and percentage of the demographic information are summarized in Table 3.1.
Second, Pearson correlation analysis was also conducted for the pilot test. The result of correlations analysis indicated that all the main variables were significantly and positively correlated as shown in Table 3.2. Psychological contract breach was positively and significantly related to intention to leave (r=.50, p<.01). In addition,
plateau (r=.26, p<.0.1) and the two sub-dimensions of career plateau (r=.25, r=.10, p<.01). Career plateau was positively and significantly related to intention to leave (r=.28, p<.01). Also, the two sub-dimensions of career plateau were positively and significantly related to intention to leave (r=.21, r=.19, p<.01).
Third, the Cronbach’s alpha values of pilot test are shown that the career plateau was 0.72 and the sub-dimensions of career plateau were 0.77 and 0.75; psychological contract breach was 0.85 and the intention to leave was 0.80.
Table 3.1.
Descriptive statistics for pilot test (n=43)
Frequency Percentage --- Item Frequency Percentage
Gender Education
Table 3.2. Mean, standard deviations, correlations, and reliability for pilot test (n=43)
Notes. Numbers in parentheses represent Cronbach’s alpha value. **p < .01 *p < .05
Data Analysis
This was a quantitative study being conducted using questionnaires. Data collected as part of this study were tabulated and entered into the Statistical Package for the Social Sciences (SPSS) 19.0 version and AMOS 18.0 for Windows. The analysis method included Cronbach’s alpha, descriptive statistics, Pearson correlation analysis, confirmatory factor analysis,and hierarchical regression analysis:
Cronbach’s Alpha
The reliability was tested by internal consistency and stability. Reliability is a concept that determines the precision of measurements. The degree of precision indication by Cronbach’s alpha coefficient indicates the stability and consistency of measurements. The higher Cronbach’s alpha coefficient means a higher consistency of each dimension. According to Guieford (1965) and Nunally and Berstein (1994), if the Cronbach’s alpha value was over 0.7, it was reliable and acceptable. The reliability analysis using Cronbach’s alpha was conducted to determine the reliability of all scales. The Cronbach’s alpha of psychological contract breach in this study was .86. The Cronbach’s alpha of career plateau serving as a moderator was .71. For the Cronbach’s alpha of the two sub-dimensions of career plateau were .76 and .71.
Moreover, the Cronbach’s alpha of intention to leave was .88.
Descriptive Statistics
This study adopted descriptive statistics (means, standard deviations, percentages and frequencies) to describe the characteristics of participants. The frequency distribution and percentage number were used to identify the participants’ distribution pattern. Hence the descriptive statistics in this study included the results of demographic information which contained gender, age, marital status, level of education, job tenure, job position and job title.
Pearson Correlation Analysis
To understand the correlation among the career plateau, psychological contract breach and intention to leave, Pearson coefficient were used to examine the correlations. A correlation of +1 means that there is a perfect positive linear relationship between variables; on the contrary, a correlation of -1 means that there is a perfect negative linear relationship between variables. A correlation below 0.4 indicates low correlation. A correlation of 0.4 to 0.7 indicates medium correlation. A correlation above 0.7 indicates high correlation.
Confirmatory Factor Analysis (CFA)
The researcher conducted confirmatory factor analysis (CFA) to evaluate the distinctiveness of the measures used in this study by using AMOS 18.0 before testing the hypotheses. When the value of chi square dividedly degree of freedom was below 3, it implied a good model fit. Moreover, the whole model fit was also examined by other fit indices which were the root-mean-square error of approximation (RMSEA), the comparative fit index (CFI), the normed fit index (NFI), the incremental fit index (IFI), the Tucker-Lewis non-normed index (TLI), and goodness-of-fit index (GFI).
When RMSEA was below .08, it implied a good model fit (Browne & Cudeck, 1993), and when CFI, NFI, IFI, TLI and GFI scores were above 0.90, it implied a good
model fit (Bentler & Bonett, 1980; Carmines & McIver, 1981).
Hierarchical Regression Analysis
Hierarchical regression analysis was used to examine the relationships among psychological contract breach, career plateau, and intention to leave. The researcher used this method to examine the relationships among variables and its sub-dimensions.
Firstly, since all independent and dependable variables were continuous variables;
hence, hierarchical regression was used to examine hypotheses in this study to understand the independent variable’s (psychological contract breach) impact on dependent variable (intention to leave). Based on Baron and Kenny’s research (1986), the moderation effect was tested by hierarchical regression analysis.
This study conducted three steps to examine the moderator effects of career plateau on the relationship between psychological contract breach and intention to leave. Firstly, control variables were entered into SPSS. Secondly, psychological contract breach and career plateau were entered to test the two variables’ impact on the intention to leave. Lastly, multiplication item of psychological contract breach and career plateau was entered to examine their interaction effect on intention to leave.