This chapter covered the research framework, research hypotheses, research design, measurement, and data collection and analysis methods. This research adopted quantitative approaches to investigate the relationship among workplace spirituality and the job involvement under M&A circumstances. In addition, this study also checked the moderating effect on the perceived organizational support.
Research Framework
The independent variable is workplace spirituality. The dependent variable is job involvement. Perceived organizational support serves as the moderator to examine the relationship between workplace spirituality and job involvement. The research framework was shown below, see Figure 3.1.
Figure 3.1. Research framework
Research Hypotheses
According to the previous literatures, the research purpose and the research questions, the hypotheses were developed below.
20
Hypothesis 1. Workplace spirituality has a positive effect on job involvement.
Hypothesis 2. The relationship between workplace spirituality and Job involvement is moderated by perceived organizational support.
Samples and Data Collection
The target sample in this study was all employees in the companies that have been through Merger and Acquisition process. The target sample was not limited in certain industry as long as he/she fulfilled the requirement of being through a certain period of time under M&A process and was still working in the companies that adopted M&A strategy.
After the financial crisis, companies struggle to find a way to bring the victory back again. Many business strategies have been used and M&A strategy is one of the business approaches that have been wildly adopted. But those business moves affect the employees’
right. To cut down the cost, companies lay off some of the employees and downsized the original organization. The phenomenon M&A brought caused some of the strong psychological reactions from the employees and even aroused the protest from the labor union. Therefore, this study chose this specific group of people to conduct the study in order to investigate the relationship within the chosen variables.
Paper-based questionnaire and online survey were used to collect as much data as possible to use in this study. In the beginning of the questionnaire, the respondents were asked whether they have been in the M&A process or not to identified the status of the respondents. Also in this study the HR personnel has been approached to deliver more questionnaires to the target sample in the companies.
In this study, the snowball sampling strategy was adopted. Snowball sampling strategy was done by consulted with people who is acquaintance with the person who is working in a company undergoing the M&A process or has been through the process.
21
The total distribution number of paper based questionnaires was 550 and 406 questionnaires returned back. The response rate is 74%, and 248 respondents (45%) fit the requirement of the study which were the employees who had M&A experiences and still working in the remain company. Moreover, 104 responses were received through online survey and total 33 fit the requirement of the study. The total amounts of valid responses were 281.
Measurement
This study adopted well-developed measurement scales to measure all the variables and the precision of measurements were confirmed through the reliability analysis and use Cronbach’s alpha coefficient (Cronbach α) to show the “stability” and “consistency” of the instruments. The higher value of Cronbach α implies the better consistency of the instrument.
The acceptable value is higher than 0.7 (Nunnally&Berstein, 1994).
Workplace Spirituality
Workplace spirituality refers to one’s sense of meaning and purpose in their work and the connectedness to another person or their workplace community (Ashmos & Duchon, 2000). The scale was developed by Ashmos and Duchon (2000) and Milliman et al. (2003).In this study, we selected three dimensions, totally 21 items in the scale. The measures were designed on a 7-point Likert scale, ranged from 1 (strongly disagree) to 7 (strongly agree).
The high score indicates that the person is highly agreed with descriptions of the items under the M&A process. The Cronbach’s alpha for workplace spirituality was 0.97. As below, the source and Cronbach’s alpha for each of the measures were described.
Meaningful work.
Meaningful work is a fundamental aspect of workplace spirituality which involves having a deep sense of meaning and purpose in one’s work (Milliman et al., 2003). In this dimension, six items were adopted from Ashmos and Duchon (2000). Example item included
22
“The work I do is connected to what I think is important in life.” The Cronbach’s alpha in this study was 0.92.
Sense of community.
Sense of community implies to the belief that employees see themselves are deeper connected with others in their workplace, feel supportive and are able to express their feeling freely (Miller, 1992; Milliman et al., 2003). Seven items were developed by Milliman et al.
(2003) to focuses on the sense of community experienced by workers. Example item was “I believe people support each other.” The Cronbach’s alpha in this study was 0.93.
Alignment with organizational values.
Alignment with organizational values indicates the employees believe their organization is sharing the same or similar purpose and value with them (Ashmos & Duchon, 2000). There are eight items in this dimension, adopted from Ashmos and Duchon (2000) spirituality scale.
The example item was “I feel connected with the organization’s goals.” The Cronbach’s alpha in this study was 0.97.
Job Involvement
Job involvement indicated to the degree that individuals psychologically identify with their present job (Kanungo, 1982). For highly involved employees, their jobs seem to be more important and they tend to develop more interest in it (Mudrack, 2004)..The scale was developed by Kanungo (1982) which contains 10 items and designed on a 5-points Likert scale, ranging from 1(strongly disagree) to 5 (strongly agree). The high score of the scale implied that people have higher level of job involvement. The example item included “I am very much involved personally in my job.” The Cronbach’s alpha in this study was 0.90.
Perceived Organizational Support
Perceived organizational support indicate that the employees form a general perception
23
concerning the degree to which the organization values their contributions and cares about their well-being (Eisenberger et al., 1986). The scale was developed by Eisenberger et al.(1986) which contains 36 items. In this study, the short version from Eisenberger et al.
(1997) which contains 8 items was used. The example item was “Help is available from my organization when I have a problem.” The Cronbach’s alpha was 0.90 in this study.
Respondents indicate their agreement with each item on a 7-point Likert scale, ranging from 1 (strongly disagree) to 7 (strongly agree).
Control Variables
The literatures revealed several demographic variables that had possibly influence on job involvement. Hence, in this study those demographic variables were put in the questionnaire as personal information in order to get the data to clarify the relationship between dependent and independent variables by controlling the influence of demographic variables.
Job position
The person who works in the higher position (manager) has indicated to develop higher job involvement (Mckelvely &Sekaran, 1977). The respondents were asked to provide their job position during the M&A process happened. Answers was coded as 1 (general employee), 2 (junior manager), 3 (mid-level manager), 4 (senior manager), and 5 (others).
Company type
The level of job involvement will change with different type of organizations (Brown, 1996). The respondents were asked to identify their company before M&A. The company type was coded as 1 (domestic company), and 2 (foreign company).
Salary
It has been proved there were positive correlation between salary and job involvement, indicated that better salary payment cause stronger job involvement (Ruh, White, & Wood,
24
1975). The annual salary was asked with different set of range and coded as 1 (Under 500,000 NTD), 2 (500,001 NTD ~ 700,000 NTD), 3(700,001 NTD ~ 800,000 NTD), 4 (800,000 NTD ~ 1,000,000 NTD), 5(Above 1,000,000 NTD).
Questionnaire Design
The questionnaire of this study was generated by several well-developed measurement scales with valid items that match each target variables. Each scale was originally written in English. Considering the target population are Taiwanese in order to make sure the target participants can understand each items. In this study, the questionnaire was translated into Chinese.
Because in this study all the scales were put in one questionnaire, each of the variables was answered at the same time from the same person, this might cause common method variance (CMV) situations. In this study, some methods were applied to avoid the CMV problem.
Based on the suggestion mentioned in the behavioral research: A critical review of the literature and recommended remedies (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003), the research purpose and the variables in this study were concealed in the questionnaire. Also, this study adopted different point Likert scale for different variables in the questionnaire. For instance, workplace spirituality adopted 7-point Likert scale while job involvement adopted 5-point Likert scale. To prevent the respondents from filling the questionnaire randomly in the study, the reverse items were also being adopted in the questionnaire.
Translation and Back Translation
Since the measurement scales this study adopted were all written in English, in order to fit target sample of this study. The questionnaire was translated into the language that the participants are familiar with, which is Chinese. The self-translated items were compared with the published translated items and consulted with the expert to ensure the items could
25
well represent this study. In order to ensure the translation of Chinese has the consistent meaning with the original English version. Back translation was conducted in this study.
The main purpose of doing back translation is to make sure the translated questionnaire contains the same meaning with the original questionnaire. Therefore, an MBA student who is a native speaker of English and fluent in Chinese was approached to assist in translating the question items back into English. The consistency of the questionnaire was ensured through adopting this technique.
Pilot Test
The purpose of conducting pilot test was to confirm the reliability of the measurement and to examine the relationship between all the variables before the big amount of the questionnaires were distributed. The questionnaires were first being translated into Chinese, to be able to let the target sample be able to response the questions.
First, the descriptive statistics analysis was applied to understand the demographic information about the samples. It indicates that most of the respondents were female (69.1 %).
Most of them were aged from 36 to 55 (66 %). As for the educational background, 89 % of the respondents got the bachelor degree. Most of the respondents were general employees (83.6%) and the total tenure is around 1 to 5 years (61.8%).
Secondly, the reliability and Pearson correlation analysis were also conducted for the pilot test to confirm the reliability of the instruments and briefly observed the relationship among each variable. The mean, standard deviation, correlations, and reliability for pilot test have been examined. The Cronbach’s alpha above 0.7 is acceptable (Nunnally&Berstein, 1994). In this study the Cronbach’s alpha for each instrument were: 0.98 for workplace spirituality, 0.92 for job involvement and 0.87 for perceived organizational support. The result of the correlation analysis indicated that workplace spirituality and job involvement were significantly positive correlated (r=.67, p<.001).
26
Data Analysis
This quantitative study was constructed by delivered and analyzed the questionnaire.
The statistical software of Statistical Package for the Social Sciences (SPSS) v21.0 and AMOS v22.0 were used to analyze the data collected from the target samples. The whole analysis methods included reliability and validity, descriptive statistics, Pearson correlation analysis confirmatory factor analysis, and hierarchical regression analysis.
Cronbach’s Alpha
To examine the internal consistency and stability, the reliability was conducted in the study. Reliability is a concept to determine the precision of the measurement. Use Cronbach’s alpha coefficient to show the stability and consistency of measurements in order to present the degree of precision. The higher value of Cronbach’s alpha indicates a higher consistency of each dimension (Nunnally & Berstein, 1994). When the Cronbach’s alpha value was over 0.7, it is reliable and acceptable. The reliability analysis using Cronbach’s alpha was conducted to determine the reliability of all scales. The Cronbach’s alpha of workplace spirituality in this study was .98. The Cronbach’s alpha of perceived organizational support serving as a moderator was .90. Cronbach’s alpha of job involvement was .90.
Descriptive Statistics
This study adopted descriptive statistics which included percentages and frequency distribution to describe the characteristics of participants. The percentage and frequency report were used to identify the participants’ distribution pattern in gender, age, educational level, company type, position, salary, and tenure.
Pearson Correlation Analysis
The correlation analysis was conducted in this study in order to investigate the direction and strength of linear relationship between the independent variables and dependent variables.
With the result, it can then explain the relationship between workplace spirituality is
27
positively or negatively related to job involvement. It is important to note that a high correlation coefficient may not necessarily imply multicollinearity existed between the variables. Since the data were confirmed by CFA, the constructs investigated in the study are also confirmed as distinct constructs. Hence, if CFA is confirmed and the correlation coefficient value is still high, it means that the variables are highly related, and the value of regression can still be accepted. Correlations analysis was tested by using SPSS 20.0 software.
Confirmatory Factor Analysis (CFA)
The Confirmatory factor analysis (CFA) were conducted to examine the validity of the questionnaire by using AMOS 22 before the proposed hypotheses has been tested, in order to confirm whether the data suit a hypothesized measurement model that based on theory or previous researches. In this study we adopted indicators that had been proposed in previous study to examine the construct validity of the workplace spirituality, job involvement and POS. The indexes were presented as below:
Chi-square (χ²) test was to determine the overall model fit and showed the difference between expected and observed covariance matrices. The value lower than 3 implied a good model fit (Jöreskog & Sörbom, 1993) . Except the chi-square test there were still many other statistical analysis to test the goodness of fit of the experimental value and theoretical value.
Root Mean Square error of approximation (RMSEA) used as an additional statistics to determine fit in the large sample sizes. If the value of RMSEA equals to 0 indicates a perfect model fit. The value range from 0.05 to 0.1 indicates a medium model fit and the value that is higher than 0.1 indicates a bad model fit (Browne, Cudeck, Bollen, & Long, 1993). Goodness of Fit Index (GFI) higher than 0.8 represents the observed data fit in the experimental distribution well (Carmines & McIver, 1981) . Normed Fit Index (NFI) shows the ratio of the difference in the value for the model and the independent model. Comparative Fit Index (CFI)
28
and Tuckers-Lewis index (TLI) = Non Norm Fit Index (NNFI) account the sample size in based on NFI. CFI, NNFI, and NFI were higher than 0.9, which then represent desirable indexes (Bentler & Bonett, 1980; Tucker & Lewis, 1973) . To conclude all, the above indexed range from 0 to 1. To be precisely, the value close to 1 represents a better model fit. The summarized fit indexes were shown as below in Table 3.1
Table 3.1
Summary of Fit Indexes
Fit Indexes Threshold References
Chi-Square/ df <2 or 3 Jöreskog and Sörbom (1993)
RMSEA <0.08 Browne et al. (1993);
GFI >0.9 or >0.8 Carmines and McIver (1981)
NFI >0.9 Bentler and Bonett (1980)
CFI >0.9 Bentler and Bonett (1980)
TLI/NNFI >0.9 Tucker and Lewis (1973)
In this study, the original 21 items of workplace spirituality was divided into three parcels and the fit indexes were presented as follows: Chi-square (χ2)= 643.96,
RMSEA=0.10, CFI=.93, NFI=.90, IFI=.93, TLI=.92, GFI=.82. The chi-square (χ2) of POS was 55.27, it presented good fit indexes while it’s RMSEA=.83, and the rest of the indexes were: CFI=.98, NFI=.97, IFI=.98, TLI=.97, GFI=.95. And as for the chi-square (χ2) of job involvement was 144.05 .The other fit indexes of job involvement were: RMSEA=.11, CFI=.93, NFI=.91, IFI=.93, TLI=.90, GFI=.90. The results of CFA for each variable were shown in Table 3.2.
29
Table 3.2.
Results of Confirmatory Factor Analysis (n=281)
χ2 df χ2/df RMSEA CFI NFI IFI TLI GFI
WS 643.96 183 3.51 .10 .93 .90 .93 .92 .82
POS 55.27 19 2.90 .83 .98 .97 .98 .97 .95
JI 144.05 32 4.50 .11 .93 .91 .93 .90 .90
3-factor model 2669.9 696 3.84 .10 .80 .76 .80 .80 .61 Notes. RMSEA = Root Mean Square Error of Approximation; CFI = Comparative Fit Index;
NFI = Normed Fit Index; IFI = Incremental Fit Index; TLI = Taker–Lewis Index, GFI = Goodness of Fit Index. WS represents workplace spirituality, POS represents perceived organizational support, and JI represents Job involvement.
Hierarchical Regression Analysis
Since both of the dependent and independent variables were continuous variables. Baron and Kenny (1986) suggested the moderation effect can be tested with hierarchical regression analysis. Three steps were conducted to examine the moderation effect of perceived organizational support in this study. First we entered the control variables, following up to enter the workplace spirituality and perceived organizational support simultaneously to test the two variables’ impact on the dependent variable. Before calculated the interaction terms of workplace spirituality and perceived organizational support, those two variables were entered by subtracting the mean from the original ones to reduce the multicollinearity problem (Aiken & West, 1991). The last step was to enter the multiplication of workplace spirituality and perceived organizational support to test the interaction effect on job involvement.
30
31