This chapter provided the results based on the proposed hypotheses in this research.
Moreover, this study showed all of the findings based on descriptive statistics, and correlation analysis, the confirmatory factor analysis (CFA), and structural equation modeling (SEM) to test hypotheses were reported and discussed. Also, the part of the discussions of this research provided the explanation and interpretation of the results.
Descriptive Statistics Results Sample Characteristics
The descriptive statistics helps to analyze the overview of the sample characteristic including frequency and percentage. A total of 196 valid responses (186 responses are employees and 15 responses are managers) were collected and analyzed in these descriptive statistics results.
In this part, there are two tables which divided into descriptive statistics for managers and descriptive statistics for employees. In addition, this type of analysis is composed of gender, age, job tenure at the current company, and education level. Table 4.1. shows the results for managers.
The majority of the managers were male (10) at 66.7% and only 33.3% were female managers (5).
Most of the managers were very aged with 73.3% at the age of 51 years old or above, 20% at the age of 41-50 years old, and 6.7% at the age of 31-40 years old. For the job tenure, most of the managers have been working in the organization for 15 years or above, accounting for 93.3%, and only 6.7% for 11-15 years. For the education level, 9 managers completed their Master degree accounting for 60% and 40% represented those managers who completed Bachelor degree.
46 Table 4.1.
Descriptive Statistics for Managers (N=15)
Variables Description Frequency Percentage
Gender Male Note. N = Number of participants
The characteristics of the employees are presented in table 4.2. It shows the results that the majority of those employees were male (102) at 54.8% and female (84) at 45.2%. For the age of the participants, it can be divided into four groups, the most populous age group with nearly 73 respondents is the age of 31-40 years old accounting for 39.2%. There is also a larger number of participants with ages of 23-30 years old (53) with 28.5%, 51 years old or above with 17.3%, and 41-50 years old with 15.1% respectively in the age of employees. For the job tenure in which the participants have been working in, 71 participants accounting for 38.2% have been working in the organization for 1-5 years, 51 participants with a total of 27.4% have been working in the organization for 6-10 years, 46 participants with a total of 24.7% have been working in the organization for 15 years or above. Results also show that only 18 participants accounting for 9.7%
have been working in the organization for 11-15 years. For the education level, the majority of respondents completed their Bachelor degree (132) accounting for 71%, 19.9% for Master degree, and 9.1% for an Associate degree.
47 Table 4.2.
Descriptive Statistics for Employees (N=186)
Variables Description Frequency Percentage
Gender Male Note. N = Number of participants
According to the correlation analysis, the results show that gender, age, and job tenure have no relationship or association with those three variables, so it contrasts with Mumford and Gustafson (1988), and Wang et al. (2014) which they suggested that gender, age, and job tenure may impact on creativity of those employees’ behavior. However, the results found that education level has a positive significance with those three variables as Simonton (1984) mentioned. So, the findings suggest that education level could predict employee’s outcome and competence especially creative performance because when people have a proper education, they know how to apply their knowledge to improve creativity and innovation.
48
The Relationships among the Variables
For the greater understanding of the relationship between variables, and the correlation analysis were implemented. Table 4.3. shows the results of means, standard deviation, and correlation of LMX, intrinsic motivation, and employee creativity. Moreover, the table also shows gender, age, job tenure, and education level as a control variable. The partial correlation (r) indicates that gender has a non-significant influence in controlling for the relationship among LMX, intrinsic motivation, and employee creativity (r = .013, p > .05; r = -.059, p > .05; r = -.142, p > .05). In the same way, the results of table 4.3. show that age as a control variable has a non-significant results with those three main variables (r = -.041, p > .05; r = -.090, p > .05; r = .025, p > .05)., also the same as job tenure, it has a non-significant results to those three main variables(r
= .018, p > .05; r = .128, p > .05; r = .106, p > .05). For the education level, table 4.3. shows the partial correlation results, there are a positively significant correlations among LMX, intrinsic motivation, and employee creativity while controlling for education level. These results indicate that education level had a little influence in controlling for the relationship between LMX and employee creativity (r = .291, p < .01; r = .262, p < .01; r = .298, p < .01). In addition, the correlation results in table 4.3. illustrate the relationship between LMX and intrinsic motivation, also the relationship between LMX and employee creativity. And it was found that LMX has a positive significant relationship with two variables intrinsic motivation (r = .406, p < .01) and employee creativity (r = .539, p < .01). Intrinsic motivation has a positive correlated with employee creativity (r = .618, p < .01).
49 Table 4.3.
Mean, Standard Deviations and Correlations (N = 186)
Variables Mean S.D. 1 2 3 4 5 6 7
Note. N = Number of participants, **p < .01
Therefore, the results in this part show that gender, age, and job tenure have no significance to those three variables. On the other hand, the results discovered that education level has a positive significance with those three variables as Simonton (1984) mentioned. So, the findings suggest that education level may predict employee’s outcome and their ability, especially in creativity performance. Moreover, the results show that LMX has positive significant relationships with intrinsic motivation and employee creativity.
The Associations among LMX, Employee Creativity, and Intrinsic Motivation
In this study, path analysis was conducted to test the hypotheses, model and investigated the direct effect, indirect effect, and total effect among three variables (LMX, intrinsic motivation, and employee creativity)as a part of SEM.In this part, the method of processing for path analysis by Baron and Kenny (1986) was adopted to investigate whether the intrinsic motivation was a mediator variable between LMX and employee creativity. The procedures for path analyses were followed through a requirement below:
1. The path regression between LMX and employee creativity must be significant.
2. The path regression between LMX and intrinsic motivation must be significant.
3. The path regression between LMX and intrinsic motivation on employee creativity must be significant.
50
According to Baron and Kenny (1986), after three-step above was met the requirement, the multiple regression was conducted to examine the direct effect, indirect effect, and total effect of the three variables. If the direct effect of LMX on employee creativity was significant and less than its total effect when intrinsic motivation is controlled, then intrinsic motivation is a partial mediator. On the other hand, if the direct effect of LMX on employee creativity was non-significantwhen intrinsic motivation is controlled, then intrinsic motivation is a full mediator.
Hypothesis 1: LMX quality is positively related to employee creativity. Table 4.4. and table 4.5. shows the path analysis results for the variables. LMX is significantly and positively related to employee creativity (β = .40, p < .01); thus, the results supported hypothesis 1. And it is confirmed to the first step of Baron and Kenny.
Based on the initial results, LMX has a substantial association with employee creativity.
So, it supports hypothesis 1 which mentioned that LMX is positively related to employee creativity. As LMX theory (Graen & Uhl-Bien, 1995) noted, high-quality LMX relationship tend to be more interactive between leader and subordinate, they rely on each other because when employees perceived the positive relationship with their leader, they acquire more support, resources, materials, and guidance from leaders which is resulted in creative outcome. Moreover, the results of the analysis are consistent with many previous studies (Aleksić et al., 2017;
Carnevale et al., 2017; Gu et al., 2018; Janssen & Van Yperen, 2004; Qu et al., 2017), they argued that the positive LMX relationship (high-quality) allows employee to access the resources, materials, and direction for supporting and increasing their creativity. Additionally, the results are in line with Tierney et al. (1999), which supported the relationship between LMX and employee creativity and it was measured by two-source perspective (rated by leader and employee).
Therefore, since subordinates perceived a positive relationship with their leaders, they tend to be more creative to work in the organization.
Hypothesis 2: LMX quality is positively related to intrinsic motivation. Table 4.5.
shows that LMX quality is significantly and positively related to intrinsic motivation (β= .56, p
< .01); so, hypothesis 2 is supported. And it is reached to the second step of Baron and Kenny test.
As a result, hypothesis 2 (LMX is positively related to intrinsic motivation) is supported.
The results highlight LMX has a strongly positive relationship with intrinsic motivation. So, the results of the analysis are consistent with self-determination theory (SDT) (Ryan & Deci, 2000), it mentions that the relatedness is the main component of intrinsic motivation when subordinates
51
interact with their leaders or managers for building a feeling of trust, they seem to be more intrinsically motivated. In addition, the finding is in line with previous studies (Chaudhry et al., 2012; Jensen & Bro, 2018; Rahbi et al., 2017), it suggested that leaders should provide a high-quality relationship with their subordinates for improving subordinate’s outcome and motivation in the workplace.
Hypothesis 3: Intrinsic motivation is positively related to employee creativity.Table 4.4. and table 4.5. show that intrinsic motivationis significantly and positively related to employee creativity (β= .47, p < .01); so, hypothesis 3 is supported.
Hypothesis 3 explored the relationship between intrinsic motivation and employee creativity. And the findings support that intrinsic motivation has a substantial effect on employee creativity. As Dweck (1986) noted, creativity is occurred when a person who is doing his or her job with commitment and efficiency to create creativity. It implied that intrinsically motivated person seems to interest and enjoy with his or her task which increasing creative idea. Moreover, the findings align with the SDT theory (Ryan & Deci, 2000), it mentions that intrinsically motivated person seems to be more spontaneous to learn and apply his or her current knowledge to establish the creativity. Additionally, the results are concurrent with many studies (Hannam &
Narayan, 2015; Muñoz-Doyague et al., 2008; Shalley & Perry-Smith, 2001; Shalley et al., 2004;
Zhang et al., 2015), they argued that intrinsic motivation enhances creativity because intrinsically motivated person more likely to embrace positive thinking, cognitive autonomy, and challenging.
These factors are the consequence of creativity.
Hypothesis 4: Intrinsic motivation has mediated the relationship between LMX and employee creativity. Table 4.5. shows that LMX has a positive significance on intrinsic motivation (β= .56, p < .01), and also intrinsic motivation has a positive significance on employee creativity (β= .47, p < .01), it implied that intrinsic motivation is a mediator; then, hypothesis 4 is supported. And the third step of Baron and Kenny's test is confirmed. In addition, the direct effect of LMX (β= .40, p < .01) on employee creativity is positive significance and less than its total effect (β= .66, p < .01) when intrinsic motivation is controlled. It implied that intrinsic motivation is the partial mediation between LMX and employee creativity.
Structural equation modeling used path analysis to estimate a collection of relevant relationships among variables concurrently (Baron & Kenny, 1986). Figure 4.1. shows that the
52
partial mediation model is a suitable model (χ2= 146.263 (df = 135, p = .240), χ2/df = 1.083, GFI
= .926, AGFI = .896, RMR = .046, RMSEA = .021, CFI = .995).
According to Gu et al. (2018), it confirmed that LMX cannot directly be related to employee creativity without mediating effect, it should have a mediator or moderator inserts between these two variables. The findings in this study found that intrinsic motivation is the partial mediation between LMX and employee creativity. It implied that LMX can directly be related to employee creativity, but it not much stronger than apply with intrinsic motivation. So, LMX enhances intrinsic motivation, and then intrinsic motivation further effect on employee creativity.
It is clear that intrinsic motivation is necessary for employee creativity. The findings align with Amabile (1983); Zhang and Bartol (2010), it suggested that the relationship between leader and subordinate can increase employees’ intrinsic motivation to improve and develop their outcome especially creativity in the workplace. Similarly, many previous studies (Buch et al., 2014; Pan et al., 2012; Schermuly, Meyer, & Dämmer, 2013) supported that LMX is positively related to employee creativity through intrinsic motivation. Moreover, the results are contrary to Buch et al.
(2014), it mentioned that high-quality LMX is negative related to intrinsically motivated person because an intrinsically motivated person do not request for extra motivation to increase performance outcome. However, the findings in this study found that high-quality LMX is positive related to intrinsic motivation because when subordinate has a good relationship with the leader, they perceive relatedness to create a basis of intrinsic motivation.
Table 4.4.
Results of the Associations among LMX, Employee Creativity and Intrinsic Motivation Variables Employee Creativity Intrinsic Motivation
Note. DE = Direct effects. IE = Indirect effects. TE = Total effects. **p < .01.
53
Note. β = Beta. SE = Estimate of the standard error of the covariance. t = t-statistic. R² = Squared multiple correlations. <--> = Do not report because of constrained parameters. **p < .01.
54 Figure 4.1. Structural equation modeling (path analysis)
55
Summary
This study investigated the relationship between LMX and employee creativity within the Ministry of Education in Thailand. Also, it examined the mediating role of intrinsic motivation on the relationship of LMX-employee creativity. Table 4.13. shows all four hypotheses are supported by the results of this study.
Table 4.6.
Summary of the Results
Hypothesis Results
H1 LMX quality is positively related to employee creativity. Accepted H2 LMX quality is positively related to intrinsic motivation. Accepted H3 Intrinsic motivation is positively affected on employee creativity. Accepted H4 Intrinsic Motivation is mediated the relationship between LMX and
Employee Creativity.
Accepted
Based on the findings and discussions in this research, the association between LMX and employee creativity is accepted, also the intrinsic motivation as a mediator between LMX and employee creativity is accepted. Therefore, the accepted results show that each of the hypothesis is proved in this research.
56
57