Moderation of Nationality on Work Design, Person-Job

From table 16, step 3 or total effect moderation model of Edwards and Lambert (2007) showed that after including P-J fit, MPS, two interaction terms into the model, the results showed both interaction terms are significant. The interaction of P-J fit and nationality has a significant interaction to predicted job satisfaction (t = -3.15, p < 0.01). Surprisingly, the interaction of P-J fit and nationality is not

significant in table 9 (t = -1.89, n.s.), but became significant in table 16. The b coefficient changed from -0.143 (n.s.) to -0.254 (p < 0.01) as shown in figure 16.

Thus, the interaction of P-J fit and nationality moderated the relationship between P-J fit and job satisfaction.

However, the interaction of MPS and nationality still significant (t = 4.45, p < 0.01), which means that MPS has a strong impact on job satisfaction.

Moreover, after adding all the variables into the model, figure 16 showed that the b coefficient changed from 0.007 (p < 0.01) to 0.011 (p < 0.01).

For Thais, P-J fit can increase more effect on job satisfaction because the result showed that there is a stronger relationship of P-J fit towards job satisfaction for Thai people. Even though the researcher had increased another interaction into the model, but the interaction of MPS and nationality still strongly predicted job satisfaction.

In summary, Thais are stronger than Taiwanese because whenever Thais have more P-J fit, they are going to have higher job satisfaction. On the contrary, Taiwanese are stronger than Thais because whenever Taiwanese people have more MPS, they are going to have higher job satisfaction.

4.3.3.4.2 Moderation of Nationality on Work Design, Person-Job fit and Job Performance

In accordance with table 17, step 3 total effect moderation model of Edwards and Lambert (2007) showed that after included P-J fit, MPS, two interaction terms into the model, the results showed that only the interaction of P-J fit and Nationality is significant.

Thereafter, the interaction of P-J fit and nationality has a significant interaction to predicted job performance (t = - 4.62, p < 0.01). The b coefficient was -0.294, which indicated that the effect of interaction of P-J fit and nationality on predicting job performance is less for Taiwan sample. For Thais, the interaction of P-J fit and nationality can increase more effect on job performance.

According to figure 17, it indicated that after including all variables into the

model, the interaction of P-J fit and nationality became more significant, the b coefficient changed from -0.270 (p < 0.01) to -0.294 (p < 0.01). Unfortunately, for

the interaction of MPS and nationality, even the b coefficient changed from 0.000 (n.s.) to 0.003 (n.s.), but the interaction still insignificant.

It is vital to note that when there is a strong P-J fit, Thai people are going to perform better on job.

Figure 16: The Moderation of Nationality on MPS, Person-job Fit and Job Satisfaction

Note: The number is the b coefficient; (a) is second stage of Edwards and Lambert (2007); (b) is direct effect of Edwards and Lambert (2007); (c) is total effect of Edwards and Lambert (2007); ** p < 0.01

Figure 17: The Moderation of Nationality on MPS, Person-job Fit and Job Performance

Note: The number is the b coefficient; (a) is second stage of Edwards and Lambert (2007); (b) is direct effect of Edwards and Lambert (2007); (c) is total effect of Edwards and Lambert (2007); **p < 0.01

!69 Table 15: Parameter Estimation and Model Summary of Moderation of MPS, P-J Fit and Nationality (Thai n=204, Taiwanese n=1509)Job Satisfaction (Dependent Variable) Step 1 . Step 2 . Step 3 . bbetat bbetat bbetat Constant2.44527.12**2.78516.31**0.8423.07**Control Variables Gender 0.1030.0592.49**0.1020.0582.38**0.0870.0492.25** Age0.0070.0582.43**0.0070.0632.46**0.0070.0592.54**Independent Variables MPS0.0160.43318.11**0.0100.2674.38**0.0010.0180.30** Nationality-0.406-0.160-2.69**0.4480.1771.62** P-J Fit 0.6510.6298.51**Interactions MPS*Nat.0.0070.2192.97**0.0110.3334.45** P-J Fit*Nat.-0.254-0.411-3.15**R 20.206**0.211**0.353**F124.281**76.691**111.781**p< 0.001**< 0.001**< 0.001**

R 20.005**0.142**

F change 4.419**157.695**p0.012**< 0.001**

Note: MPS is Motivating potential score; Nat. is Nationality; Code 0 = Thai data; Code 1 = Taiwanese data; ** p < 0.01; * p < 0.05

70 Table 16: Parameter Estimation and Model Summary of Moderation of MPS, P-J Fit and Nationality (Thai n=204, Taiwanese n=1509)Job Performance (Dependent Variable) Step 1 . Step 2 . Step 3 . bbetat bbetat bbetat Constant3.53545.22**3.54725.28**2.34110.69**Control Variables Gender -0.123-0.099-3.25**-0.121-0.097-3.13**-0.119-0.096-3.22** Age-0.005-0.069-2.22**-0.006-0.076-2.03**-0.007-0.090-2.50**Independent Variables MPS0.0090.37912.23**0.0100.3895.98**0.0040.1602.26** Nationality0.0040.0030.04**0.9460.6384.40** P-J Fit 0.4130.6387.02**Interactions MPS*Nat.0.000-0.017-0.20**0.0030.1251.38** P-J Fit*Nat. -0.294-0.789-4.62**R 20.144**0.144**0.206**F52.659**31.555**34.767**p< 0.001**< 0.001**< 0.001**

R 20.000**0.062**

F change 0.058**36.789**p0.943**< 0.001**

Note: MPS is Motivating potential score; Nat. is Nationality; Code 0 = Thai data; Code 1 = Taiwanese data; * p < 0.05; ** p < 0.01

Chapter 5

Discussion and Conclusions

Why people are willing to work? Is it because of money, benefit or something else? This study assumed that people are willing to work because they want to do something that fit their need. And this study also assumed that people are willing to work because work design is rational. Thence, this study was conducted in order to find out answers of the question “Why people are willing to work?” In this study, the researcher had included three things together, which are work design, P-J fit and work outcomes.

Work design is a rational model, which can make people able to follow accordingly. Even though all these issues had been done in other countries before, but the researcher believed that it does not only happen in western culture but also happen in Thailand and Taiwan. Thus, it is curious for the researcher to test these issues in Thailand, as well as in Taiwan, as the researcher believed that there should have differences in people lives and job. For instance, in Thailand, people want to have happy life and do not want to have much stress on job, which is difference from other countries.

Person-job fit is also important towards work because researcher assumed that when people feel fit with their job, they will be happy and perform their best on work assignment, which is refers to job satisfaction and job performance included in this study.

In this study, online survey and traditional pencil-paper questionnaire were carried out. Questionnaires were distributed and collected within two months. A total of 204 valid samples were collected. The data were obtained by using personal relationships of the researcher to invite their friends, relatives and co-workers to respond to the questionnaire, samples include employees from different fields. The reliability and validity testing showed that the structure of work design, job satisfaction and job performance was good.

5.1 Summary of Results of Hypothesis Testing

After analyzed the data, the following table showed on the summary of hypothesis testing result.

72 Table 15: Summary of Results of HypothesesNo.HypothesesResultsH1Work design is positively related to work outcomes. H1aTask characteristics are positively related to job satisfaction. SupportedH1bTask characteristics are positively related to job performance. SupportedH1cKnowledge characteristics are positively related to job satisfaction. Partially SupportedH1dKnowledge characteristics are positively related to job performance. Partially SupportedH1eSocial characteristics are positively related to job satisfaction. SupportedH1fSocial characteristics are positively related to job performance. Partially SupportedH1gWork context is positively related to job satisfaction. SupportedH1hWork context is positively related to job performance. Partially SupportedH2Work design is positively related to person-job fit. H2aTask characteristics are positively related to person-job fit. SupportedH2bKnowledge characteristics are positively related to person-job fit. Partially SupportedH2cSocial characteristics are positively related to person-job fit. Partially SupportedH2dWork context is positively related to person-job fit. SupportedH3Person-job fit is positively related to work outcomes. H3aPerson-job fit is positively related to job satisfaction.SupportedH3bPerson-job fit is positively related to job performance. SupportedH4Person-job fit will mediate the relationship between work design and work outcomes. H4aPerson-job fit will mediate the relationship between task characteristics and job satisfaction.SupportedH4bPerson-job fit will mediate the relationship between task characteristics and job performance. SupportedH4cPerson-job fit will mediate the relationship between knowledge characteristics and job satisfaction.SupportedH4dPerson-job fit will mediate the relationship between knowledge characteristics and job performance. SupportedH4ePerson-job fit will mediate the relationship between social characteristics and job satisfaction. SupportedH4fPerson-job fit will mediate the relationship between social characteristics and job performance. SupportedH4gPerson-job fit will mediate the relationship between work context and job satisfaction.SupportedH4hPerson-job fit will mediate the relationship between work context and job performance. SupportedH5Nationality will moderate the relationship between work design, person-job fit and work outcomes. Partially Supported

According to hypothesis 1, the researcher assumed that work design is positively related to work outcomes. While work design includes four dimensions, which are task characteristics, knowledge characteristics, social characteristics and work context. The result showed that most sub dimensions of work design are positively related to job satisfaction and job performance. But, job complexity is not positively related to job satisfaction and job performance. Fortunately, this is similar to Taiwan data, therefore the researcher suggested that future research should try to expand more on this dimension. Besides, received interdependence and equipment use are not related to job performance. In conclusion, task characteristics are positively related to job satisfaction and job performance, so the hypothesis 1a and hypothesis 1b are supported. But knowledge characteristics in hypothesis 1c and hypothesis 1d are partially supported. For social characteristics and work context, both of them are positively related to job satisfaction, which can support hypothesis 1e and 1g; but are partially supported hypothesis 1f and 1h.

From hypothesis 2, the researcher assumed that work design is positively related to person-job fit. And the result showed that the sub dimensions of work design are positively related to person-job fit, except for job complexity and received interdependence. In summary, task characteristics and work context are positively related to person-job fit, so the hypothesis 2a and 2d are supported. However, some of the dimension in knowledge characteristics and social characteristics are not positively related to person-job fit, which means the hypothesis 2b and 2c are partially supported.

In accordance with hypothesis 3, the researcher assumed that person-job fit is positively related to work outcomes. Under this condition, the result supported that person-job fit is positively related to job satisfaction and also job performance. This can support hypothesis 3a and 3b.

Baron & Kenny (1986) quoted that mediation illustrated the generative mechanism of the independent variable able to influence on dependent variable, it means that mediation used to explain on how or why external physical events take on internal psychological significance. To clarify, there are three variable systems in mediation model, which have direct impact of the independent variable to outcome variable, the impact of mediator to outcome variable, and path from independent variable to mediator.

When a variable reaches the following conditions, it operates as a mediator:

(a) variations in the independent variable essentially account for variations in the presumed mediator, (b) variations in the mediator essentially account for variations in the dependent variable, and (c) when path from mediator to outcome variable and path from independent variable to mediator are controlled, it makes a previously significant relationship of the independent and dependent variables no longer significant (Baron & Kenny, 1986). The strongest demonstration of mediation occurs when path from independent variable to outcome variable becomes zero, it is a strong evidence of a single or dominant mediator, which is known as fully mediation. On the contrary, if it does not become zero, this indicated that there are multiple mediating

factors, which is known as partially mediation (Baron & Kenny, 1986; Derek, Kristopher, Zakary & Richard, 2011).

In agreement with hypothesis 4, the researcher assumed that person-job fit will mediate the relationship between work design and work outcomes. The result showed that person-job fit can fully mediate the relationship between sub dimensions of work design and job satisfaction; except for decision-making autonomy, work methods autonomy, information processing, specialization, social support, ergonomics and work condition because their previous relationship still significance, it means that these dimensions are partially mediated. Furthermore, person-job fit can fully mediate the relationship between sub dimensions of work design and job performance;

excluding task variety, task identity, information processing, social support and feedback from others because they still have direct effect on job performance, which means that these few dimensions are partially mediated. So, to sum it up, even some sub dimensions of work design are fully mediated and some of them are partially mediated. But it appeared that all of them supported hypothesis 4 which assumed that person-job fit will mediate the relationship between work design and work outcomes.

In short, the result supported hypothesis 4a, 4b, 4c, 4d, 4e, 4f, 4g and 4h.

Moderator is a variable that affects on strength or direction of relationship between independent variable and dependent variable. This variable affects zero-order correlation between two variables, which means that it occurs when the direction of the correlation changes (Edwards & Lambert, 2007).

Under hypothesis 5, the researcher assumed that nationality will moderate the relationship between work design, person-job fit and work outcomes. Besides, in this study, the researcher would use motivating potential score instead of work design, as the outcome of motivating potential score is similar to work design. Surprisingly, some of them support the hypothesis, while some of them do not support hypothesis.

Moreover, differences way of moderation occurred because under some condition Taiwanese have a stronger effect, and under some conditions Thais have a stronger effect. In this study, the result showed that nationality moderates the relationship between P-J fit and job performance. Moreover, nationality moderates the relationship between MPS and job satisfaction. After all, the researcher had included MPS, P-J fit and the two interactions into the model, and the results indicated that nationality moderates the relationship between MPS, P-J fit, job satisfaction and job performance. In accordance with these results, hypothesis 5 is partially supported.

To sum up, result can support almost all of the hypotheses. However, there were only few hypotheses that cannot be fully supported by the result. What matters here is work design, P-J fit and work outcomes are important factor of work. It is also important to keep in mind that nationality is also important because different countries have differences work conditions as shown in the result of the hypothesis 5.

5.2 Summary of the Findings

5.2.1 Work Design is a Strong Predictor of Person-Job Fit and Work Outcomes

Results demonstrated that work design has a positive influence on person-job fit, except for job complexity and received interdependence. And the results also showed that work design mostly has a positive influence on work outcomes, which are job satisfaction and job performance. Previous researches indicated that people want to have autonomy, get social support, training, etc. in order to be satisfied in job (Chompoopong, 2010; Nookoonungaree, 2012). In this study, it supported once again that good work design can effect on job satisfaction.

In work design, it included all kinds of work that everyone has to face. For instance, how to do your work, what kind of knowledge you should have, is there anyone who willing support you, how do you feel about your working environment, etc. From this study, it can prove that work design is very important, because it is not only effect on fit, but can also effect to work outcomes, which is very important towards every work.

Moreover, work design is not only important to Thais and Taiwanese but also important to other countries too, as work design study has been done in many countries; such as, US, Germany, Poland, Spain, Dutch, France, Indonesia, Italy, Portugal, Russia, etc. Fortunately, this study can once again prove the important of work design.

It comes as no surprise that work design is very important because it can effect on fit of people towards job and also work outcomes. And this can be proved not only in this study, but also in other studies that had been conducted before.

5.2.2 Person-Job Fit is an Important Mediator

As mentioned in the previous section that the strongest demonstration of mediation occurring when path from independent variable to outcome variable becomes zero, there is only a single mediator left, which means that it can fully mediate. However, if it does not become zero, this indicated that there are multiple mediating factors, which known as partially mediation (Baron & Kenny, 1986; Derek, Kristopher, Zakary & Richard, 2011).

The results in chapter four proved that P-J fit has a significant positive influence on work outcomes, which are job satisfaction and job performance. This means that P-J fit is very important because it can moderate the relationship between work design and work outcomes.

In order to favor work outcomes, it is important to understand the antecedents of person-job fit, which is the same as Li and Huang (2010). Moreover, Arora (2000) quoted it would be best if company could find individual who has the same requirements as a particular job need. As we can see from the result, it is no wonder

that why company should find people who have same requirements of job. In conclusion, person-job fit is important as it can directly effect on work outcomes.

5.2.3 The Moderation of Culture Differences is Complicated

Culture is values, perceptions, beliefs and behaviors that can be affected by family, relatives and others in the same environment. Culture can be differences in different countries.

Thailand and Taiwan are both Asian countries, some of the norms have an influence from traditional Chinese culture. For instance, people do not want to loose face in public and do not want to confront with each other, etc.

It is well known that Thai people want to live their life with happiness, which is similar to Hofstede finding. Therefore, they often choose job that can fit their need and want. And if they enjoy with their need, it also effect on their satisfaction and also performance. Thai people often go to work late and leave punctually. Nevertheless, there were different magnitudes of correlation between variables across nationality.

Amazingly, there are two ways of moderation that have been found in this study.

On one hand is a stronger relationship of Thai people. First of all, when people feel fit they are willing to perform better job. Thence, Thai slope get steeper than Taiwanese 0.270, when there is a change in one unit. Secondly, people can feel more satisfy, if they feel fit with their job, which is the result of good work design. When there is one unit change, the slope of Thai is going to be steeper 0.254. Last but not least, the result showed that when there is a good work design, people would feel more fit on their job and finally result in better job performance. Therefore, when there is one unit change, the slope of Thai is going to be steeper than Taiwanese 0.294.

Psychologically, in Thailand, if there is a good work design, people will feel more fit, which finally result in satisfaction and best performance.

On the other hand, there is also a stronger relationship of Taiwanese than Thais. The result showed that when MPS is going higher, people feel more satisfy on their job. So, whenever there is one unit change, the slope of Taiwanese is going to be steeper than Thai people 0.011. In this point, researcher should pay more attention to work design because the direct effect towards job satisfaction is so strong for Taiwanese. To crown it all, when there is a good work design, it can make people feel more satisfied in their job.

In conclusion, even though there are similarities among different cultures, but differences in culture still exist as a result showed in this model.

5.2.4 The Differentiation of Effect of Outcome

According to the result, the two majors of outcomes have different effect. For instance, job satisfaction is more concerned on psychological thinking, which is about feeling of each person. In contrast, job performance is more concerned on an action,

According to the result, the two majors of outcomes have different effect. For instance, job satisfaction is more concerned on psychological thinking, which is about feeling of each person. In contrast, job performance is more concerned on an action,