Chapter Overview
This chapter outlines the methodology. The chapter contains the Research Framework, Hypothesis, research procedure, instrumentation (measurement instrument), reliability and validity of instruments and data analysis. The framework will display the way variables are being tested, and the section Research Process will explain step by step how the research has been lead. All the details concerning the method, the procedure and the instrumentation are provided.
Research Framework
This research framework was developed in accordance with the literature reviews. From reviews, it was noticed that the integration into Taiwanese culture of foreign workers has a positive relation with their commitment at work, thus with their performance and success (Mowday et al., 1982)
36 Figure 3.1. Research general framework.
Hypotheses
Based on the literature review, the following null-hypotheses were created:
H1 Work Satisfaction has a positive effect on employees Organizational Commitment.
H2 Cross-Cultural Adaptation has a positive effect on employees Organizational Commitment.
H3 Foreign Worker Characteristics have a positive effect on their Organizational Commitment.
H4 Commitment of foreign employees in Taiwan has a positive effect on their Employee’s Performance.
Research Procedure
The research procedure is divided into eight stages, shown in the figure below. At the beginning of this study, the researcher must clearly identify the problem and issue. After this step, the search for relevant prior literatures on the subject to study has been effected. With a lot of researches and data from the past, a theoretical framework has been developed. Once the frame of the research has been determined, the researcher has developed a questionnaire related to the subjects of study. Finally, the researcher will focus on the conclusion of his study, and leave an open part with questions for future researchers.
Identifying the subject of the study Literature review
Develop the framework of the study Develop the questionnaire of the study
Apply the questionnaire to the subject Data collection and analysis Conclusion and suggestions
Conclude the thesis
37 Figure 3.2. Research process.
Measurement Instrument
The instrument consisted of 5 variables with total amount of 72 questions. After the pilot test, a statistical analysis was initially conducted to see the validity of the instrument. Some items were dropped as either the items could not pass the requirements of reliability test or factor analysis test.
The choice of the researcher in the data collection instrument is a composite questionnaire adapted from several models obtained from the literature review. It was distributed among the target population through paper based copies and online questionnaire. The questionnaire utilizes an ordinal scale or Likert scale with responses categorized from 1 = completely disagree, 2 = disagree, 3 = neutral, 4= agree, 5 = agree completely.
The measurement instrument consists of 5 variables with a total of 72 questions in addition to a section for demographic questions. Part I of the questionnaire contains Work Satisfaction (WS);
Part II contains Cross-Cultural Adaptation (CCA); Part III contains Foreign Worker Characteristics (FWC); Part IV contains Organizational Commitment (OC) ; Part V contains Employee’s Performance (EP) with a total of 15 research sub dimensions. All constructs in the questionnaire were adapted from pre-validated measures in existing correlated researches. Furthermore, peer reviews and expert reviews were utilized to keep intact the validity of the instrument used in the questionnaire. Pilot test was initially conducted to ensure the validity of each question, before gathering all the necessary data for the study.
Work Satisfaction (WS)
In the Part I Work Satisfaction, the 15 questions were divided into 3 variables, adopted from various resources. The dimension Compensation (COM) was adopted from Maister (2008), the dimension Culture (CUL) was adopted from Belias and Koustelios (2004), and the dimension Perspective of Evolution (PER) was built from the study of Farrell (2013).
Cross-Cultural Adaptation (CCA)
In the Part II Cross-Cultural Adaptation, the 15 questions were divided into 3 variables. The validity and reliability of the 15 items were also being tested in the pilot test. The three variables measured in this section are: Social Integration (SI), Adaptation to Lifestyle (AL), Language Proficiency (LP), and they have all been adapted from the study of Kuo and Roysircar (2006).
38
Foreign Worker Characteristics (FWC)
In the Part III Characteristic of the Foreign Worker, the 10 questions were adopted from Shalom and Schwartz, (2005). The validity and reliability of the 10 items were also being tested in the pilot test. The two variables measured in this section are: Values (VAL) and Personality (PER).
Organizational Commitment (OC)
In the Part IV Commitment of the foreign worker, the 17 questions were adopted from Allan and Meyer (1990). The validity and reliability of the 17 items were also being tested in the pilot test. The three variables measured in this section are: Organizational Commitment (OC) (17 questions), three variables are measured: Affective Commitment (AC), Normative Commitment (NC) and Continuance Commitment (CC).
Employee’s Performance (EP)
In the Part V Performance of the foreign worker, the 15 questions were divided into 3 variables: Attitude and Stress (AS), Productivity (PRO), and Group Cohesion (GC). The dimension Attitude and Stress was adopted from Kuo and Roysicar, (2006), the dimension Productivity was adopted from Kolesnikova (2012). The dimension Group Cohesion was adopted from Eys and Lougheed (2009).
Demographics Data
This section contained 4 nominal items of demographic information such as: participants’
Gender, age and length of time working in the organization (tenure). All constructs were adapted from pre-validated measures in existing researches.
Sample and Participant Selection
In order for the study to attain its desired outcome, the researcher used a quantitative method approach for data analysis. The research subjects of this study are foreign workers who are currently having a job in Taiwan for at least 1 year. This population was selected because these workers are more likely to meet and encounter difficulty of integration and stabilization in their work life rather than foreign workers already settled in Taiwan for more than 10 years. The population feedback
39
desired for the formal study is a total of about 210 foreign workers. They were all willing to answer the survey in order to implement the study successfully.
The questionnaire has been delivered within the organizations directly towards foreign workers. A pilot test is of a desired population of 45 has also been made previously.
The researcher has met with people through his personal network in Taiwan and with the help of local contacts in that he has accumulated during the years he has been settled in Taiwan.
Therefore, the data collection is the result of a convenient sampling, because the researcher didn’t select in a random way the respondents. This is a part that the researcher suggests that could be improve in further study, to provide a result as objective as possible. As the researcher already had several different experience of work in Taipei, he went through his former colleagues and former managers as a way to distribute the questionnaires and collect more data. His database was mostly found in a recruitment company he was working in, and where he had an access to numerous resources and respondents for the sake of this study. He has received the permission from the company owning the database in order to utilize it and gather enough resources for the research.
Validity and Reliability of Instrument Face Validity
In this research, the reliability of these items were exanimated once again through a pilot test.
This pilot test was effected in order to assed to validity of the research and allow the researcher to continue his research on a broader scale. It was carried out based on a population of 45 foreign worker respondents. After this, a main study was effected in order to develop and give a more accurate result of the study. The reliability of instrument is the external and internal consistency of measurement. The scales used in the questionnaire should be evaluated again because items are from different research and each research has different samples. The validity of an instrument is the degree to which scales measure what researchers claim they measure (Williams & Monge, 2001).
As respondents are foreign workers, to obtain a greater level of understanding of the questions from them, the questionnaire was sent in English Language. The researcher has assessed that the huge majority of respondent has a sufficient English proficiency to respond to the survey. Each item has been reviewed by a native English speaker in order to make the questions understandable and clear to workers from any nationality. Also, the instrument was basically gathered in English from numerous and various works and studies previously carried out on a similar subject. It was then put altogether for the sake of this research and the researcher has provided the list of these sources.
40 Table 3.1.
Reliability of Instruments Reliability
Type Theoretical Meaning Reliability of Instrument
Reliability Analysis
The degree to which a scale consistently reflects the construct (Field, 2007) others, especially in the scientific community (Neuman, 2011).
The questionnaire should examined by native speakers.
The questionnaire was also initially undergone pilot tests.
Construct Validity
It is a type of measurement validity that uses multiple indicators and has two subtypes:
how well the indicators of one construct converge or how well the indicators of different constructs diverge (Neuman, 2011).
Convergent validity is established by using Explanatory Factor Analysis (EFA). Before the EFA test, KMO and Bartlett’s test for Sphericity are conducted. If the items are confirmed then EFA can be conducted.
Divergent validity is established by using Confirmatory Factor Analysis (CFA).
41
Validity and Reliability
The researcher has lead a pilot study in order to evaluate the individual reliability of each item and the outer loadings of the measurements. The vast majority of indicators had outer loadings above than the required 0.5. This means that every items with a factor loading superior to 0.5 in the questionnaire have relations with the items they are supposed to measure. Same as for the main study, the pilot has shown values of Cronbach Alpha higher than 0.70. The items for the variable of Work Satisfaction (15 items) has a value of 0.856 and its composite reliability is 0.785; Cross-Cultural Adaptation (15 items) reaches a value of 0.857 and its composite reliability is 0.779; for the variable of Foreign Worker Characteristics (10 items) the value of Cronbach’s Alpha is 0.860 and its composite reliability is 0.740; for the variable of Organizational Commitment (17 items), the value reached 0.189 and its composite reliability of 0.321; for the variable Employee’s Performance, the value reached 0.768 and the composite reliability reached 0.722. The values can be observed in table 3.2.
CCA Cross-Cultural Adaptation 0.779 0.857
FWS Foreign Worker Characteristics 0.740 0.860
OC internal consistency, reliability and internal validity. Indeed, each variable are above the required value of 0.70. Also, all these variables are higher than this amount for the composite reliability.
Putting the focus on the variable Work Satisfaction, the researcher has observed that the dominant factor is Compensation, and the weakest one is Culture. For the variable of Cross-Cultural Adaptation, the dominant sub variable is Language Proficiency and the weakest one is Social Integration. Focusing on the variable of Foreign Workers Characteristics, the most dominant sub variable is Personality, and the weakest one is Values. Within the variable Organizational
42
Commitment, we can observe that the most dominant variable is Affective Commitment, and the weakest one is Continuance Commitment. Focusing on the variable Employee’s Performance, the dominant sub variable is Productivity, and the weakest sub variable is Attitude and Stress.
Concerning the factor loading, we can observe that 12 questions of the variable Organizational Commitment are below the acceptable level of 0.5, and could be eliminated or improved. The items that the researcher has decided to drop are displayed in table 3.3.
Table 3.3.
Pilot Outer Loading of all items via PLS (N=45) (Pilot Test-Before Dropping Items) Constructs Number of Note. AC= Affective Commitment; NC= Normative Commitment; CC= Continuance
Commitment
The researcher has decided here to drop some item from the variable Organizational Commitment only, and to keep the other item. This decision has been made because following the
43
results of the Pilot Test, the variable Organizational Commitment is the only one presenting a problem and not reaching the required 0.7 value of Cronbach Alpha. In the table 3.3, we can see that only 5 item on 17 has been kept and considered as reliable by the researcher. All the dropped item are going to be replaced by new one in the main study, in order to obtain a result with a present constructs have different factor loading as other constructs. For this reason, the requirements are not met. After dropping items, the value of Cronbach alpha increased significantly from .321 to .758. This value was affected before since the organizational commitment values are too different from each other, since we are dealing with foreign workers. Their differences are too high and this affects the consistency of the answers hence it is reflected in a low Cronbach alpha value.
Note. AC= Affective Commitment; NC= Normative Commitment; CC= Continuance Commitment
The improvement from .321 to .758 can be considered as important. Therefore the researcher has decided to apply it in order to improve the reliability of the model as a whole as shown in the table 3.4. The table 3.5 displays the PLS findings for the Pilot test and shows that Work
44
Satisfaction has a strong positive effect on Organizational Commitment, the null hypothesis is thus accepted. Cross-Cultural Adaptation has a weak positive effect on Organizational Commitment.
Therefore, this null hypothesis is rejected. Foreign Worker Characteristics also have a weak positive effect on Organizational Commitment. Organizational Commitment has a strong positive relationship with Employee’s Performance therefore the null hypothesis is accepted. Table 3.5 in the appendix summarizes the research results for the pilot test.
Table 3.5.
PLS Path Analysis Results (Pilot Test =45)
Path Hypothesis β-path Adj. t-value Sig. Direction Null Hypotheses
WS → OC H1 0.231 5.063 *** + Accepted
CCA → OC H2 0.198 0.436 * + Rejected
FWC → OC H3 0.301 0.887 * + Rejected
CC → EP H4 0.366 6.877 *** + Accepted
Note. * p < .1, ** p < .05, *** p < .001.
WS= Work Satisfaction; CCA= Cross-Cultural Adaptation, FWC= Foreign Worker Characteristics, OC= Organizational Commitment, EP = Employee’s Performance.
Although the pilot study (n=45) has previously confirmed that the instrument was in part valid, the researcher has lead one more test of validity and reliability through a main study, comprising 210 respondents. As said previously, this second test was lead in order to improve the validity and reliability of the instrument.
In the aim of assessing a correct reliability for an acceptable result, the researcher gathered with success the values of Cronbach’s Alpha and composite reliability. To be appropriate, factor analysis has to be over the value .70. In order to assess the validity of the measurement instrument, both convergent and discriminant validity will be measured. Composite reliability and average variance will also help to take the measure of the convergent validity and reliability of the measurement instrument. As it was n the pilot study, the minimum value recommended for Cronbach’s Alpha is .70.
The table 3.6 shows that all the values of Cronbach’s Alpha are higher than minimum value of .70, which indicates that all the values are accepted, after some items were deleted from the
45
Cross-Cultural Adaptation dimension. In the table below we can observe the construct of Work Satisfaction with 15 items has a Cronbach alpha value of 0.856; Cross-Cultural Adaptation with 15 items has a value of 0.857; Foreign Worker Characteristics with 10 items has a value of 0.860;
Organizational Commitment with 17 items has a value of 0.758 and Employee’s Performance with 15 items has a value of 0.768. All the measurement has been carried out in the same way as they were in the pilot study. Only the item dropped by the researcher can explain the different of results between the pilot study and the main study.
Table 3.6.
Cronbach’s Alpha Results for all Dimensions (n=210)
Constructs Number of Items Cronbach Alpha
Work Satisfaction 15 0.856
Cross-Cultural Adaptation 15 0.857
Foreign Worker Characteristics 10 0.860 Organizational Commitment
In order to assess the validity of the measurement instrument, the researcher took the measure of convergent validity and discriminant validity. Composite reliability and average variance extracted were measured in order to examine the convergent validity and reliability of the measurement instrument. There is a minimum value that is recommended for composite reliability.
This value is .70 (Wong, 2013). As all the values of composite reliability are higher than minimum value, this shows that the values are accepted. In the table 3.7 below, we can observe the construct of Work Satisfaction has a composite reliability value of 0.881; Cross-Cultural Adaptation has a value of 0.880; Foreign Worker Characteristics has a value of 0.813; Organizational Commitment has a value of 0.813, and Employee’s Performance has a value of 0.817.
As for the Average Variance Extracted (AVE), it was determined in order to verify the degree of convergence validity for each of the variables in this researcher. The AVE values are generally meant to be higher than 0.5. The Table 3.7 below shows that each of dimension converges well. It also indicates that most of the values are rather higher than the minimum requirement, except for one variable that is Employee’s Performance. This variable indicates a value of (0.485). Yet, even under 0.5, this value can still be accepted because it is still regarded as sufficient. The conclusion
46
that all the constructs met convergent validity can be made.
Table 3.7.
CCA Cross-Cultural Adaptation 0.880 0.857 0.526
FWC Foreign Worker Characteristics 0.888 0.860 0.719
COM
According to Henseler, Ringle and Sarstedt (2015), it is still uncertain to determine the exact threshold level of the Heterotrait-Monotrait Ratio (HTMT), and it still can be debated. Meanwhile some other authors have been suggesting a threshold of 0.85 (Clark, Georgellis, & Sanfey, 1998).
There even are some others proposing a higher a value of 0.90 (Gold, 2001). For this reason, when the HTMT value doesn’t reach the value 0.90, discriminant validity has been established between two reflective constructs. On the table 3.8 below, results bring the following conclusion : there is discriminant validity between work satisfaction and cross-cultural adaptation (HTMT=0.490), followed by foreign characteristics and work satisfaction (HTMT=0.433), cross-cultural adaptation and foreign worker characteristics (HTMT=0.555), work satisfaction and organizational commitment (HTMT=0.618), cross-cultural adaptation and organizational commitment (HTMT=0.668), work satisfaction and Employee’s Performance (HTMT=0.485), cross-cultural adaptation and Employee’s Performance (HTMT=0.446) and organizational commitment with Employee’s Performance (HTMT=0.693). The relation foreign worker characteristics and organizational commitment (HTMT=0.707) is slightly higher but still within the acceptable range.
Moreover, the relation foreign worker characteristics and Employee’s Performance (HTMT=0.216) is the lowest, establishing the strongest discriminant validity.
47 Table 3.8.
Heterotrait-Monotrait Ratio (HTMT) Results for all Dimension (n =210)
Constructs WS CCA FWC OC EP
WS - - - -
CCA 0.490 - - -
FWC 0.433 0.555 - - -
OC 0.618 0.668 0.707 = -
EP 0.485 0.446 0.421 0.693 -
Note. WS= Work Satisfaction; CCA= Cross-Cultural Adaptation, FWC= Foreign Worker Characteristics, OC= Organizational Commitment, EP= Employee’s Performance
Table 3.9, reports the Fornell and Larcker discriminant validity results for all dimensions.
AVE and shared variance estimates should be compared to assess discriminant validity the “square root” of AVE of each latent variable should be greater than the correlations the latent variables. The table below demonstrates all the average variances extracted from the study are higher than the correlations shown below them or to their left, which shows good discriminant validity.
Table 3.9.
Fornell-Larcker Criterion Results for all Dimension (n =210)
Constructs WS CCA FWC OC EP
WS 0.577 - - -
CCA 0.421 0.577 - - -
FWC 0.388 0.487 0.667 - -
OC 0.541 0.550 0.618 0.461 -
EP 0.406 0.326 0.334 0.547 0.487
Note. WS= Work Satisfaction; CCA= Cross-Cultural Adaptation, FWC= Foreign Worker Characteristics, OC= Organizational Commitment, EP= Employee’s Performance.
48
49