Chapter 2: Literature Review
2.5 Hypothesis
2.5.5 Moderation Effect of Nationality on Work Design, Person-Job Fit and
As work design is made up of all kinds of work that should use people who feel fit with those things to perform job in order to get a good outcome.
The researcher thought that psychology could get an influence from physiology. People are born in many different environments and circumstances. One of the differences is nationality. When individuals are born in different countries, they have different nationality according to place where he or she born. And this issue would also be brought into focus in this research. Nationality refers to an ethnic group.
In different country, people tend to have different belief, norm, culture, etc. These factors would influence on different belief or action people take. Therefore, research beliefs that nationality would moderate on work design, person-job fit and work outcomes.
For example, Taiwanese pay more attention on pragmatic. Thence, Taiwanese are thriftiness, save and invest for future life. This is main reason why Taiwanese people try to work hard. So, the relationship between work design, person-job fit and work outcomes should be more significant. But Thailand is more normative. People do not pay much attention on their future. Therefore, this study predicted that the relationship between work design, person-job fit and work outcomes is not as strong as Taiwanese.
Hypothesis 5: Nationality will moderate the relationship between work design, person-job fit and work outcomes.
Chapter 3
Research Methodology
This chapter reveals into four main sections. First section discourses about sample. Secondly, this chapter mentions about instruments. Third, this section discusses on procedure of this study. Finally, this chapter also manifests an analysis in this study.
3.1 Sample
The sample uses in this study consisted of two parts. First part of sample is a pilot study. Second part of sample is a formal study.
3.1.1 Pilot study
In the pilot study, there were 35 participants to fill up the questionnaire in order to make sure that the questionnaire can work well with all people. The people who fill up the questionnaire must work on full-time job and can give feedback about the questionnaire. Therefore, the researcher sent questionnaire to friends and relatives who still working and having a full-time job. All participants in the pilot study are relatives and friends in different companies and/or different department, having different working conditions. The researcher directly contacted each participant by using social media; such as Facebook and Line. After they finished the questionnaire, participants were asked to give their opinion about the questionnaire and difficulty of answering questionnaire.
3.1.2 Formal study
In the formal study, the researcher used an online fill up form and pencil-paper to collect data. Within 2 months, there were 220 participants who fill up this questionnaire.
The sampling method of this study was using personal connection and asking people to help distribute the questionnaire. As the researcher had worked in Thailand before, therefore the researcher had connection with different company managers and workers in different fields. However, the researcher was not in Thailand, so the questionnaire was mainly collected by help of friends, relatives and ex-coworkers.
The questionnaire had been sent as a PDF file and link through e-mail, Facebook and Line to friends, ex-coworkers, and relatives who work in different fields and cities in Thailand, but most of the sample data are mainly from Bangkok (urban area) and sub-urban area. In Thailand, most rural people leave their hometown to work in Bangkok, therefore the data is mostly based on capital city of Thailand.
In sum, a total of 110 samples were collected online by sending Google form link to friends, relatives and ex-coworkers. After eliminating the missing data and ineffective questionnaire, there were 105 valid samples, 110 samples were collected in pencil-paper survey by sending PDF file to friends, relatives, and ex-coworkers and asked them to print out the form, but there were only 99 valid samples left. In total, there were 204 valid samples. The rate of ineffective sample was 7.27%
The participants consisted only those who have full time job and work in Thailand. The total sample consisted of 163 females (79.9%) and 41 males (20.1%).
The minimum work tenure was 1 month and the maximum was 33 years. The average age was 37.663 years old, with standard deviation of 9.366. 127 participants (62.3%) were undergraduates and 36 participants (17.6%) were postgraduates. Most people were not manager, because there were only 46 managers from the data, which included 31 basic level managerial positions and 15 high level managerial positions.
The maximum samplings were employed as secretary consisting of 35 people (17.16%), salesperson 20 people (9.80%), and accountant 18 people (8.82%). The following are type of the company; insurance company, law company, construction company, security company, architect company, spare parts company, bank, etc.
Table 1: The Sample Structure of the Formal Study
Variables Categories No. Percentage Valid
Percentage
Education Less than Junior High School Senior High School
Work design is a broader characteristic, which includes every aspect needed in a job (Morgeson & Humphrey, 2006). In past research, Morgeson & Humphrey (2006) had conducted research to check on the WDQ reliability. The results demonstrated on
excellent internal consistence reliability, which was 0.87 in average (Morgeson &
Humphrey, 2006).
Thereupon, the researcher adopted the work design questionnaire of Morgeson and Humphrey (2006). The original version of WDQ is an English version of questionnaire. But for participants’ convenience, the researcher translated work design questionnaire of Morgeson and Humphrey (2006) into Thai. The internal consistency reliability Cronbach’s alpha was 0.947.
Moreover, the researcher has also computed MPS, or known as motivating potential score. Traditionally, it comes from JCM model of Hackman and Oldham (1975). Whenever the core dimensions increase, it would also increase in MPS. On contrary, if any of the core dimensions score is low, the MPS should also low. High or low MPS can effect on motivation of an organization, which is the meaning of work and responsibility on work (Chiou, Chou & Lin, 2010).
MPS = {([(Work Scheduling Autonomy + Decision-Making Autonomy + Work Methods Autonomy) ÷ 3] + Task Variety + Task Identity) ÷ 3} × Task Significance × Feedback From Job
Table 2: Samples of Work Design Questionnaire
Task Variety 4 The job involves a great deal of task variety.
Task Significance 4 The job has a large impact on people outside the organization.
Task Identity 4 The job allows me to complete work I start.
Feedback From Job 3 The job itself provides feedback on my
Problem Solving 4 The job requires me to be creative.
Skill Variety 4 The job requires a variety of skills.
Specialization 4 The job requires a depth of knowledge and expertise.
Social Characteristics
Social Support 6 People I work with are friendly.
Initiated
Interdependence
3 Other jobs depend directly on my job.
Received
Ergonomics 3 The job involves excessive reaching (reverse scored).
Physical Demands 3 The job requires a lot of physical effort.
Work Conditions 5 The work place is free from excessive noise.
Equipment Use 3 The job involves the use of a variety of different equipment.
3.2.2 Measures of Person-Job Fit (P-J Fit)
Person-job fit is the involvement of matching person with requirements associated with a specific work (Newton & Jimmieson, 2009). It is used to match knowledge, skills and abilities of people in organization with demand or need of job (Edwards, 1991; Lawrence, 2004; O’Reilly, Chatman, & Caldwell, 1991; Sekiguchi
& Huber, 2011).
In this study, six items with a 5-point scale ranging from 1 (Strongly disagree) to 5 (Strongly agree) had been used to measure employees’ P-J fit. These items had been translated from Chiou, Wen and Liao (2008) study in order to keep equivalent for comparisons with Taiwanese data. The sample question of this study was “My work gives me a sense of accomplishment”. The Cronbach’s alpha of P-J fit was 0.900.
3.2.3 Measures of Work Outcomes
In this study, it categorized work outcomes into two types; which are job satisfaction and job performance. Therefore, this part expanded work outcomes by using the measures of job satisfaction and job performance.
3.2.3.1 Measures of Job Satisfaction
Job satisfaction is a feeling of gratefulness by doing something or experiencing something, but would dissatisfy while getting block from doing so (Locke, 1976). In this study, nine items with a 5-point scale ranging from 1 (Strongly disagree) to 5 (Strongly agree) had been used to measure employees’ satisfaction. The items were also translated from Chiou, Wen and Liao (2008) study in order to keep equivalent for comparisons with Taiwan data. The sample question was “I feel satisfy with my job”. After consider about too much questions, the researcher had chose only 3 questions as items used to measure job satisfaction. The Cronbach’s alpha of job satisfaction was 0.868.
3.2.3.2 Measures of Job Performance
Job performance refers to how well employee performs their job. It based on how employees perform their job to accomplish the goal (Devoudi & Allahyari, 2013), which is highly related to organization and individual (Sonnentag, Volmer &
Spychala, 2008).
In this study, the researcher predicts that job performance could get influence from work design and person-job fit. In this study, the researcher used 5 questions of job performance; such as, “Overall, my work quality is very good”. This study used 5-point scale ranging from 1 (strongly disagree) to 5 (strongly agree). The questions were translated from Chiou, Wen and Liao (2008) study in order to keep equivalent
for comparisons with Taiwan data. After consider about too much question, the researcher had chose only 3 questions as items used to measure job performance. The Cronbach’s alpha of job performance was 0.918.
3.2.4 Taiwanese Data
This research is going to compare the differences between different nationalities, by using the data of Thailand and Taiwan. To achieve this, the researcher uses the three data that had been collected previously in Taiwan.
The first data set is collected by Chiou, Wen and Liao (2008), which have 540 participants. In the research, the researchers had examined the reliability and validity of Traditional Chinese version of work design questionnaire (χ2 = 5193, p < 0.001, df = 2639, χ2/df = 1.97, RMSEA = 0.057, NNFI = 0.93, CFI = 0.94, SRMR = 0.071).
After eliminating invalidity data, there are only 470 valid samples that are able to compare with Thai data. The average age is 27.512, with standard deviation of 6.101.
From 470 participants, there are 269 females and 201 males. In this group, there are only 27 participants who are manager.
The second data has 266 samples, with an average age of 25.964, standard deviation of 2.991. In this sample, there are 125 female participants and 141 male participants (Chiou, Ou & Chou, in press)
The third data is collected by Chiou (2009). This data set has 773 samples, with an average age of 32.741, and standard deviation of 6.754. In this group, there are 260 female participants and 513 male participants. In this data, there are 225 people were manager, which including 222 basic level managerial positions and 3 high level managerial positions. Summing it up, the three data as described above has been used to compare with the Thai’s data.
3.3 Research Procedure
The following procedures are steps that are used in order to conduct this research. First of all is to translate questionnaire. The researcher translates the English version of the work design questionnaire that was created by Morgeson and Humphrey (2006) into a Thai version in order to ask Thai people to fill up this questionnaire. Moreover, the researcher also translates Chinese questionnaire of P-J fit, work job satisfaction and job performance of Chiou, Wen and Liao (2008) into Thai version in order to ask Thai people to fill up the questionnaire.
Secondly, the researcher asks people who are fluent in English and Thai to do a back translation into English in order to compare the meaning of each question whether the original English version and the back translate English version have the same meaning or not. After finding out the difference between two editions, the researcher edits until meaning of Thai version and English version is exactly the same. And the researcher also asks people who are fluent in Chinese and Thai to do a back translation into Chinese in order to check the meaning of each questionnaire.
After that the researcher compares and corrects mistake to make both Thai version and the original Chinese version have the same meaning.
Thirdly, the researcher conducts a pilot study by distributing the questionnaire to ask 35 people with different kinds of jobs. The most frequent problems found are length issue. Therefore, the researcher improves by deleting some of the questionnaire and dividing questionnaire into smaller sector in order to minimize the bias of participants.
Fourth, the researcher conducts a formal study by using an edited version of questionnaire. Edited version refers to the latest version of questionnaire that has been corrected of all errors occurred in pilot study. The researcher distributes the edited version of questionnaire by using online fill up form and pencil-paper form. Finally, there were 220 participants in this study.
Fifth, after all sample ready, the researcher arranges sample data, translates into English and coding all samples. Furthermore, the researcher also eliminates invalid samples. After that, the researcher does a confirmatory factor analysis to measure the questionnaire structure, in order to confirm the validity of work design questionnaire.
Sixth, the researcher has conducted a statistical analysis and analyzes the data in order to check whether the data can support this study hypothesis or not. The researcher also conducts basic analysis on the research variable for Thais, as well as for Taiwanese sample. Moreover, the researcher examines the relationship between independent variables and dependent variables. In addition, the researcher uses mediation to analyze how independents variable can influence on dependent variable.
Seventh, the researcher compares Thai’s sample result with Taiwan’s sample result in order to understand the difference in work design, person-job fit and work outcomes of people in two countries.
From above, these are steps that have been used to conduct in this research. In the end, the researcher also concludes result and discusses further about the research and future study.
3.4 Analysis
After collecting samples, the researcher analyzes the data that has been collected in order to understand whether the results can support the predictions or not.
3.4.1 Descriptive Analysis and Data Screening
After collecting data, the researcher screens the data and keypunch into computer using SPSS software, creating the data file for analysis. The mean, standard deviation, percentage, minimum and maximum of the variables data; such as age, gender, education, occupation, managerial position, along with the measures of research variables (Work design, P-J fit, Work outcomes) are listed for purpose of
checking. Finally, the correlation between variables is computed for describing the relationship between variables.
3.4.2 Analysis of Measurement Tool
In this research, confirmatory factor analysis has been used to identify factor structure of observation variables. The quality of the measurement may have an impact on the research hypothesis being tested. The CFA was performed by AMOS17. The chi-square value, CFI, RMSEA and SRMR are used to determine the adequacy of model fit to the data. The chi-square test is used for evaluating the different between expected and observed covariance matrices. If chi-square value is closed to zero, it means that there is very little different between the expected and observed covariance matrices.16 CFI has to exceed 0.90. RMSEA should be less than 0.06. SRMR should be less than 0.08 (Hu & Bentler, 1999). As a result, the well-defined factor structure is approved by the CFA.
3.4.3 Hypothesis Testing
A series of hierarchical regression analysis is performed to test on hypothesis.
In the first step, the researcher controls three variables, which are gender, age and managerial position. Secondly, independent variables are included, and moderators are tested. Third, the interaction variable is mounted.
3.4.4 Analysis of Mediation and Moderation
Mediation and moderation can be used to explain how independent variables exert its effect on the dependent variable, and when independent variables affect dependent variable and when they do not (Hayes, 2013). This study applies mediation and moderation in order to compare the strength of relationship of the model between Thailand and Taiwan.
3.4.1.1 Procedures of Mediation
Mediation is used to test how independent variables exert its effect on dependent variables. It can explain how independent variables can transmit their effect on dependent variable.
The intervention variable is known as the mediator or intermediary variable, which is a mechanism through which independent variables influence dependent variable (Hayes, 2013). The effect of independent variable on dependent variable is transmitted by mediator variable (Alwin & Hauser, 1975).
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
16!Source: Exploratory or Confirmatory Factor Analysis? By Diana D. Suhr, PhD., University of Northern Colorado.!
Researcher can tell whether a variable is a mediator when it meets four conditions. The first condition is when independent variations account for variations in a dependent variable. The second condition is when independent variations account for variations in mediator. The third condition is when mediator variation accounts for variations in a dependent variable. The last condition is when mediator adds to the model, the relationship between a dependent variable and independent variables would decrease or eliminate (Baron & Kenny, 1986; Hayes, 2013).
Figure 4: The Mediation of Work design and Work Outcomes
In this research, it examines whether person-job fit can mediate the relationship between work design and work outcomes or not. From figure 4, the researcher has to examine the mechanism resort used in the process modeling to empirically estimate and test the hypotheses in two pathways of influence of how work design could effect on work outcomes. The first pathway is the direct effect of work design on work outcomes, without passing through person-job fit. The other pathway is an indirect effect of work design on work outcome. This is to explain how work design would influence person-job fit and how person-job fit would affect the work outcome.
However, by adding the mediator, most of work design dimensions should not directly affect the work outcomes. Thence, it is useful to use mediation to explain this research.
3.4.1.2 Procedures of Moderation
Moderation is to moderate the effect of predict and outcome. Moderation occurs when the effect of independent variable on dependent variable change because of moderator (Baron & Kenny, 1986; James & Brett, 1984).
Moderation is also known as interaction. It is used when the goal is to uncover an association between two variables. The association between dependent variable and independent variables would be moderated when the size or sign depends on third variable or set of mediation variable. Moderation is used to depict moderator variables
that could influence the magnitude of the effect of dependent and independent variables (Hayes, 2013).
Moderation is conducted with the purpose of testing an interaction between the moderator and independent variables in a model of dependent variable (Hayes, 2013). It is involved in differences on individual, situation or conditions that impact on relationship of predictor and outcome (Cohen & Edwards, 1989; Taylor &
Aspinwall, 1996). But, moderation can only explain the strength or direction of the relationship between dependent variable and independent variables. The following figures have been used to explain the moderation effect (Edwards & Lambert, 2007;
Hayes, 2013).
Figure 5: First Stage Moderation Model
Edwards and Lambert (2007) explains that in accordance with figure 5, nationality acts as a moderator of the path from work design to person-job fit, which is known as the first stage of the mediated effect of work design and work outcomes.
Figure 6: Second Stage Moderation Model
According to figure 6, nationality also acts as a moderator of the path from person-job fit to work outcomes, which is known as the second stage of the mediated effect of work design and work outcomes in Edwards and Lambert (2007) theory.
Figure 7: Direct Effect Moderation Model
However, figure 7 demonstrated that nationality is a moderator of the path
However, figure 7 demonstrated that nationality is a moderator of the path