國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Chapter 4. Results and Discussion Overview
Chapter four discussed the results of this study. Furthermore, it separately discusses the results of the descriptive analysis, correlation analysis, regression analysis. This chapter aims to explain the findings and relationship among the variables. It also aims to describe variables that could be strong predictors for job acceptance intention and learning.
Descriptive Analysis
A total of 150 survey questionnaires were distributed online to targeted sample namely university and graduate school students in Taiwan who’ve had internship experiences before. Out of 151 questionnaires, 126 questionnaires were completed. After cancelling out 6 invalid questionnaires, a total of 120 questionnaires were usable, producing a final response rate of 79.4%
With regards to the gender of the respondents, 56.7% of the respondents are female.
As for the age groups of the respondents, over 80% of them are aged at 20-25 years old, with the remaining 15.8% to be from 25-30 years old. The industries that the respondents had experienced in during internship were FMCG, consulting, financial, tech, web and media.
19.33% of the respondents had experienced in the financial sector, followed by tech with 17.65% and FMCG with 13.45%.
With regards to internship duration and period, 50% of the respondents had 1-2 months of internship experience, while 25% had 3-6 months of internship experience, the remaining 11.7% and 13.3% had longer experiences from 6-9 months and 10-12 months respectively. On the other hand, 41.7% of the respondents had their internship programs during summer and winter break while the majority had their internship programs during fall
‧
undergraduate programs. The summary of the results is shown in table 4Table 4
The Demographics of the Respondents and Internship Features Demographical Variables Number of
Respondents
Internship Time Summer &
Winter Break willingness to stay in the company after the internship program. 13.3% of the respondents stated that they refused to accept a job offer from their previous employer, while 52.5% stated
‧
respondents stated that they would like to accept a job offer after the internship program. The summary of the findings is shown in table 5Table 5
The relationships of the variables were tested with the use of correlation analysis.
All the variables were significantly related except for the relationship of autonomy and job acceptance intention. The rest of the variables obtained two asterisks (p value <0.01). Thus, it can be concluded that all the variables are interrelated to each other except for autonomy and job acceptance intention. The summary of the results of the correlation analysis is shown in table 6.
‧
Clarity Mentorship Autonomy Learning Job
As illustrated in table 7, model 1 explains that three control variables, namely gender, internship duration, and degree are not significant with R squared of 0.028 only. Model 2 shows that task goal clarity significantly explained learning with R squared increased from 0.028 to 0.282. Thus, H1 is supported. H2 predicted that mentorship would be positively associated to learning. Model 3 shows a positive coefficient of (β = 0.594, p < 0.001) and an R-square of 0.364, H2 is also supported.
To test for mediation, the independent variable should be positively related to both the mediator and the dependent variable. The independent variable’s effect should disappear (full mediation) or be weakened (partial mediation) when the mediator is added to the model (Kenny, Kashy, & Bolger, 1998). When mentorship was added in model 4 of table 7 the relationship of task goal clarity and learning weaken from a coefficient of (β = 0.505, p <
0.001) to (β = 0.222, p < 0.05). Thus, mentorship partially mediated the relationship between
‧
goal clarity and learning. Mentorship had a positive coefficient of (β = 0.447, p < 0.001) with an r-square value of 0.392. H3 is partially supported.
H4 predicted that Autonomy would moderate the relationship between Task Goal Clarity and Learning. For it to be supported, the two-way Task Goal Clarity X Autonomy interaction must be both significant and positive. As shown in model 2 and model 5, task goal clarity and autonomy are both positive and significant. Thus, could significantly explained learning. However, when a two-way interaction of Task Goal Clarity X Autonomy was added to model 6, the coefficient of the two independent variables weakened. The two-way interaction had a positive coefficient of (β = 0.228, p < 0.05). Thus, H4 is supported.
The relationship of task goal clarity and learning is expected to be stronger when autonomy is high and weaker when it is low.
Table 7
Summary of Hierarchical Regression for Variables Predicting Learning (Hypothesis 1, 2, 3,4)
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
As illustrated in model 9 of table 8, H5 predicted that Learning is positively related to Job acceptance intention. Learning had a positive coefficient of (β = 0.404, p < 0.001) with an R square of 0.199. Thus, H5 is supported.
A mediation analysis was also conducted to test whether Learning mediated the relationship of Task Goal Clarity to Job Acceptance Intention. H6 predicted that learning would mediate the relationship of Task Goal Clarity and Job Acceptance Intention. As illustrated in model 10 of table 8, the relationship of Task Goal Clarity and job acceptance intention weaken from a coefficient of (β = 0.395, p < 0.001) to (β = 0.259, p < 0.01) respectively. Thus, learning partially mediated the relationship between task goal clarity and job acceptance intention. Learning had a positive coefficient of (β = 0.270, p < 0.01) with an R-square value of 0.248. H6 is partially supported.
Table 8
Summary of Hierarchical Regression for Variables Predicting Job Acceptance Intention (Hypothesis 5&6)
Step & Predictor Model 7 Model 8 Model 9 Model 10 1. Control variables
Gender -0.077 -0.080 -0.081 -0.082
Internship Duration -0.024 -0.009 -0.042 -0.026
Degree 0.177 0.161 0.11 0.122
2. Predictor variable
Task Goal Clarity 0.395*** 0.259**
3. Predictor variable (H5&H6)
Learning 0.404*** 0.270**
R-square 0.041 0.196 0.199 0.248
Adjusted R-square 0.016 0.168 0.171 0.215
*p < .05, **p < .01, ***p < .001.
‧
Chapter 5. Conclusion and Discussion