This chapter presents the results and finsings of the resaerch data. First, the descriptive statistics is presented, including the demographic characteristics of the sample group. Secondly, the Pearson correlation analyses, the linear and multiple regression analyses and t-test analyses are reported. In addition this chapter presents the discussion and conclusion of the findings.
Descriptive Statistics Sample Characteristics
A total 749 samples was selected from the 2015 data in this study. The entire sample was composed of 51.9 percent of males and 48.1 percent of females, with a mean of age of 38.44 (SD = 11.37, range = 18-76). The percent of married was 54.7, and single was 38.6. Over a quarter of the respondents (27.1) were managers at various levels. And 37.7 percent of respondents had a working spouse with full-time jobs.
A total 823 samples was selected from the 2005 data in this study. The entire sample was composed of 53.8 percent of males and 46.2 percent of females, with a mean of age of 37 (SD
= 10.08, range = 18-76). The percent of married was 61.5, and single was 34.6. Over a quarter of the respondents (28.7) were managers at various levels. And 40.8 percent of respondents had a working spouse with full-time jobs.
Genarally, the distribution of each demographic characteristics were similar, with a difference of less than 5%. And there were some interesting differences bewteen 2005 and 2015. For example, the perceptage of married in 2015 was less than 2005, even though the age distributions both in 2005 and 2015 were similar. Otherwise, it presented that people tended to get higher education compared to 2005. The results are shown in Table 4.1.
Table 4.1.
Demographic Characteristics of the Sample of TSCS 2005 & 2015
2005 (N = 823) 2015 (N = 749)
Variable Category Frequency Percentage % Frequency Percentage %
Age 18-29 234 28.4 180 24.0
Descriptive Statistic of Variables
This section presents the mean and standard deviation for each item of instrument. The research instrument has some reverse coded questions, which have been recoded in SPSS.
Descriptive statistics for flexible work arrangements. Table 4.2. illustrates the
descriptive analysis for flexible work arrangements. As can be seen from the table, the scale of flexible workhours was 3-point, the mean scores of flexible workhours in 2005 was 1.4921, while in 2015 the scores was 1.4219. The results indicated that most respondents’ worktimes were decided by their employers and they could not change them by themselves, with a lower degree of deciding workhours in 2015 than in 2005. For flexible schedule, the scale of flexible schedule was 3-point, the mean scores were 1.9915 in 2005 and 2.0320 in 2015. The results indicated that the majority of the respondents agreed with the fact that they own flexible schedule with certain limits, but the degree to organize daily schedule was improved in 2015.
Last, the scale of allowance for personal leave was 4-point, the mean scores of allowance for personal leave were 2.7230 in 2005 and 2.5861 in 2015. The results indicated the majority of the respondents believed that it was not difficult to take personal leave during working hours, however, the scores in 2015 was lower than that in 2005, which means that employees accessed this arrangement in 2015 worse than they did in 2005. The scores of each arrangement did not differ much between 2005 and 2015.
Table 4.2.
Descriptive Statistics of Flexible Work Arrangements
2005 (N = 823) 2015 (N = 749)
Code Scale M SD M SD
FWA I (Flexible Workhours) 3-point 1.4921 .66151 1.4219 .62339 FWA II (Flexible Schedule)* 3-point 1.9915 .78376 2.0320 .78865 FWA III (Personal Leave)* 4-point 2.7230 1.06529 2.5861 1.13152 Note. *represent the reverse coded items.
Descriptive statistics for job satisfaction. Table 4.3. illustrates the descriptive analysis
for job satisfaction. As can be seen from the table, the scale of item JS I was 5-point, the mean scores was 3.9392 in 2005 and was 4.0440 in 2015. The results indicated most of respondents claimed that they kept quite good relations in the workplace and were satisfied with them, and reported higher satisfaction toward their workplace relations in 2015. For the satisfaction with their main job, the scale of item JS II was 7-point, the mean scores was 5.0049 in 2005 and 5.2056 in 2015. The results showed the majority of the respondents expressed that they fairly satisfied with their jobs, and reported higher satisfaction toward their main jobs in 2015. In final, the degree of being satisfied with organization, the scale of item JS III was 5-point, the mean scores was 3.3908 in 2005 and 3.4214 in 2015. The results presented that most of respondents were neither satisfied or dissatisfied with their organizations, and reported higher satisfaction toward their organizations in 2015.
Table 4.3.
Descriptive Statistics of Jos Satisfaction
2005 (N = 823) 2015 (N = 749)
Code Scale M SD M SD
JS I (Relations)* 5-point 3.9392 .66587 4.0441 .67968
JS II (Main Job)* 7-point 5.0049 .98898 5.2056 .97037
JS III (Organization)* 5-point 3.3908 .69598 3.4215 .74432 Note. *represent the reverse coded items.
Descriptive statistics for occupational commitment. Table 4.4. illustrates the
descriptive analysis for occupational commitment. As can be seen from the table, the scale of item OC I was 5-point, the mean scores was 3.3147 in 2005 and was 3.4806 in 2015. The results indicated that most of respondents were in a neutral attitude, they felt neither easy nor difficult to find another job at least as good as their current one, and reported higher
commitment toward their current job in 2015. According to item OC II, the scale of item OC II was 4-point, the mean scores was 2.8566 in 2005 and 3.0481 in 2015. The results mean that the majority of the respondents expressed that they were unlikely to find a job with another organization within the next 12 months, and reported higher commitment toward their organizations in 2015. Last, the scale of item OC III was 4-point, the mean scores was 3.4331 in 2005 and was 3.5642 in 2015. The results showed a higher commitment that most of respondents didn’t worry about the possibility of losing job, , and reported higher job security in 2015.
Table 4.4.
Descriptive Statistics of Occupational Commitment
2005 (N = 823) 2015 (N = 749)
Code Scale M SD M SD
OC I (Job Change Difficulty) 5-point 3.3147 1.32057 3.4806 1.28180 OC II (Job Change Possibility) 4-point 2.8566 1.01634 3.0481 1.00418 OC III (Job Security) 4-point 3.4331 .81371 3.5642 .71316
Correlation Analysis
The Pearson correlation analysis was used to examine the preliminary correlation on flexible work arrangements, job satisfaction and occupational commitment. This analysis was used to measure the strength and the direction of a linear relationship between two variables.
If the correlation was high, it showed that there is a strong relationship between them. The coefficient value ranged from -1.0 to +1.0 and a value was greater than 0 indicating that there was a positive relationship and vice versa.
Correlation of TSCS 2015
Based on Table 4.5., each flexible work arrangement had a significant correlation with each dimension of job satisfaction. Flexible workhours was all positively correlated with JS I
(r = .109**, p < .01), JS II (r = .093**, p < .01) and JS III (r = .171**, p < .01) respectively.
Flexible schedule was all positively correlated with JS I (r = .164**, p < .01), JS II (r = .161**, p <.01) and JS III (r = .179**, p < .01), respectively. Allowance for personal leave was all positively correlated with JS I (r = .186**, p < .01), JS II (r = .173**, p < .01) and JS III (r
= .230**, p < .01), respectively.
However, the relationships between FWA and OC were not all significantly correlated.
Only three significant relationships were positively correlated, namely, flexible schedule and OC III (r = .130**, p < .01), allowance for personal leave and OC II (r = .086*, p < .05) and allowance for personal leave and OC III (r = .073*, p < .05). While, other relationships didn’t.
Correlation of TSCS 2005
Based on Table 4.6., each flexible work arrangement had a significant correlation with each dimension of job satisfaction. Flexible workhours were positively correlated with JS I (r
= .163**, p < .01), JS II (r = .125**, p < .01) and JS III (r = .184**, p < .01), respectively.
Flexible schedule was positively correlated with JS I (r = .140**, p < .01), JS II (r = .203**, p
< .01) and JS III (r = .166**, p < .01), respectively. Allowance for personal leave were positively correlated with JS I (r = .270**, p < .01), JS II (r = .260**, p < .01) and JS III (r
= .224**, p < .01), respectively.
However, the relationships between FWA and OC were not all significantly correlated.
Only two significant relationships were positively correlated, namely, flexible schedule and OC III (r = .087*, p < .05) and allowance for personal leave and OC III (r = .085*, p < .05).
While, other relationships didn’t.
Table 4.5.
Correlations among the Variables for TSCS 2015 (N = 749)
Measures M SD 1 2 3 4 5 6 7 8 9 10 11 12
1. Age 38.44 11.37 1
2. Education 12.79 5.90 -.430** 1
3. Management .27 .45 .154** .037 1
4. FWA I (Flexible Workhours) 1.42 .62 -.009 .125** .132** 1
5. FWA II (Flexible Schedule) 2.03 .79 .016 .172** .192** .334** 1
6. P FWA III (Personal Leave) 2.59 1.13 .004 .036 1.28** .329** .325** 1
7. JS I (Relations) 4.04 .68 .012 -.013 .100** .109** .164** .186** 1
8. JS II (Main Job) 5.21 .97 .146** -.124** .063 .093* .161** .173** .401** 1
9. JS III (Organization) 3.42 .97 .150** -.027 1.23** .171** .179** .230** .292** .410** 1
10. OC I (Job Change Difficulty) 3.48 1.28 .227** -.115** .050 -.055 -.047 .029 -.020 .127** .158** 1
11. OC II (Job Change Possibility) 3.05 1.00 .288** -.096** .058 .015 .049 .086* .163** .363** .356** .291** 1
12. OC III (Job Security) 3.56 .71 -.016 .105** .036 .071 .130** .073* .047 .025 .009 -.187** -.038 1
Note. Management: 0 = No, 1 = Yes. JS: Job satisfaction; OC: Occupational commitment. *p < .05, **p < .01.
Table 4.6.
Correlations among the Variables for TSCS 2005 (N = 823)
Measures M SD 1 2 3 4 5 6 7 8 9 10 11 12
1. Age 37.00 10.08 1
2. Education 11.45 5.64 -.154** 1
3. Management 1.71 .46 -.119** -.122** 1
4. FWA I (Flexible Workhours) 1.49 .66 .004 .098** -.215** 1
5. FWA II (Flexible Schedule) 1.99 .78 .007 .153** -.223** .376** 1
6. P FWA III (Personal Leave) 2.72 1.07 .002 -.004 -.120** .302** .289** 1
7. JS I (Relations) 3.94 .67 .027 -.003 -.048 .163** .140** .270** 1
8. JS II (Main Job) 5.00 .99 .100** .070* -.106** .125* .203** .260** .440** 1
9. JS III (Organization) 3.39 .70 .193** .003 -.164** .184** .166** .224** .379** .467** 1
10. OC I (Job Change Difficulty) 3.31 1.32 .240** -.056 .027 -.061 -.017 .011 .034 .176** .180** 1
11. OC II (Job Change Possibility) 2.86 1.01 .255** -.023 -.087* .011 .029 .032 .210** .327** .306** .294** 1
12. OC III (Job Security) 3.43 .81 -.142** .168** -.007 .044 .087* .085* .054 .073* -.004 -.245** -.027 1
Note. Management: 0 = No, 1 = Yes. JS = Job satisfaction; OC = Occupational commitment. *p < .05, **p < .01.
Hypotheses Testing
Linear regression analysis was used to analyze the relationship between FWA (flexible workhours, flexible schedule and allowance for personal leave) and JS (relations, occupation and organization), respectively.
Flexible Work Arrangements and Job Satisfaction
H1-1. Flexible workhours is positively related to job satisfaction (relations, occupation, organization). In this study, the hypothesis 1-1 was used to test whether flexible workhours was positively correlated with JS in three dimensions (relationship, occupation and organization).
According to the 2005 results, flexible workhours presented the significant effect on JS I (β = .076, p < .01) and JS III (β = .103, p < .01). The overall R2 of JS I was .005, indicating that flexible schedule explained 0.5 % of the variation in the dimension of relations. The overall R2 of JS III was .009, indicating that flexible schedule explained 0.9 % of the variation in the dimension of organization. According to the 2015 results, flexible workhours only presented the significant effect on JS III (β = .083, p < .05). The overall R2 of JS III was .006, indicating that flexible schedule explained 0.6 % of the variation in the dimension of organization.
As the results, the impact of flexible workhours was significant on JS I and JS III in 2005, but only significant on JS III in 2015. Thus, H1-1 was partially supported by the 2005 and 2015 data in the present study, and the results are shown in Table 4.7.
Table 4.7.
Linear Regression Analyses for Main Effect of Flexible Workhours on Job Satisfaction
2005 (N = 823) 2015 (N = 749)
JS β Sig. △R2 β Sig. △R2
JS I (Relations) .076* .041 .005 .026 .513 .001
JS II (Main Job) .007 .848 .000 .011 .773 .000
JS III (Organization) .103** .006 .009 .083* .032 .006 Note. Independent Variable: Flexible Workhours. Dependent Variable: JS = Job Satisfaction.
*p < .05, **p < .01.
H1-2. Flexible schedule is positively related to job satisfaction (relations, occupation, organization). In this study, the hypothesis 1-2 was used to test whether flexible schedule was positively correlated with three dimensions of JS (relationship, occupation and organization).
According to the 2005 results, flexible schedule presented the significant effect on JS II (β = .137, p<.001) and JS III (β = .078, p < .05). The overall R2 of JS II was .015, indicating that flexible schedule explained 1.5 % of the variation in the dimension of occupation variable.
The overall R2 of JS III was .005, indicating that flexible schedule explained .5 % of the variation in the dimension of organization variable. According to the 2015 results, flexible schedule presented the significant effects on JS I (β = .110, p < .01), JS II (β = .114, p < .01) and JS III (β = .095, p < .05). The overall R2 of JS I was .010, indicating that flexible schedule can explain 1 % of the variation in the dimension of relations variable. The overall R2 of JS II was .011, indicating that flexible schedule explained 1.1 % of the variation in the dimension of occupation variable. The overall R2 of JS III was .008, indicating that flexible schedule explained .8 % of the variation in the dimension of organization variable.
As the results presented in Table 4.8., the impact of flexible schedule was significant on JS II and JS III in 2005, but it was completely significant on JS in 2015. Thus, H1-2 was
partially supported by the 2005 data and fully supported by the 2015 data in the present study.
Table 4.8.
Linear Regression Analyses for Main Effect of Flexible Schedule on Job Satisfaction
2005 (N = 823) 2015 (N = 749)
JS β Sig. △R2 β Sig. △R2
JS I (Relations) .044 .236 .002 .110** .005 .010
JS II (Main Job) .137*** .000 .015 .114** .004 .011
JS III (Organization) .078* .036 .005 .095* .014 .008 Note. Independent Variable: Flexible Schedule. Dependent Variable: JS = Job Satisfaction. *p
< .05, **p < .01, ***p < .001.
H1-3. Allowance for personal leave is positively related to job satisfaction (relations, occupation, organization). In this study, the hypothesis 1-3 was used to test whether allowance for personal leave is positively correlated to with three dimensions of JS (relations, occupation and organization).
According to the 2005 results, allowance for personal leave presented the significant effects on JS I (β = .235, p < .001), JS II (β = .218, p < .001) and JS III (β = .171, p < .05). The overall R2 of JS I was .048, indicating that flexible schedule explained 4.8 % of the variation in the dimension of relations variable. The overall R2 of JS II was .042, indicating that flexible schedule explained 4.2 % of the variation in the dimension of occupation variable. The overall R2 of JS III was .025, indicating that flexible schedule explained 2.5 % of the variation in the dimension of organization variable. According to the 2015 results, allowance for personal leave presented the significant effects on the JS I (β = .142, p < .001), JS II (β = .132, p < .01) and JS III (β = .171, p < .001). The overall R2 of JS I was .017, indicating that flexible schedule explained 1.7 % of the variation in the dimension of relations variable. The overall R2 of JS II was .015, indicating that flexible schedule explained 1.5 % of the variation in the dimension of
occupation variable. The overall R2 of JS III was .025, indicating that flexible schedule explained 2.5 % of the variation in the dimension of organization variable.
As the results presented in Table 4.9., the impact of allowance for personal leave was significant on all items of JS both in 2005 and 2015. Thus, H1-3 was fully supported by the 2005 and 2015 data in the current study.
Table 4.9.
Linear Regression Analyses for Main Effect of Allowance for Personal Leave on Job Satisfaction
2005 (N = 823) 2015 (N = 749)
JS β Sig. △R2 β Sig. △R2
JS I (Relations) .235*** .000 .048 .142*** .000 .017
JS II (Main Job) .218*** .000 .042 .132** .001 .015
JS III (Organization) .171*** .000 .025 .171*** .000 .025 Note. Independent Variable: Allowance for Personal Leave. Dependent Variable: JS = Job Satisfaction. *p < .05, **p < .01, ***p < .001.
Flexible Work Arrangements and Occupational Commitment
H2-1. Flexible workhours is positively related to occupational commitment. In this study, the hypothesis 2-1 was used to test whether Flexible workhours is positively related to the three facets of OC. According to the results of 2005 and 2015, flexible workhours did not have significant effect on OC. Thus, H2-1 was not supported by the 2005 and 2015 data in the present study, the results are presented in Table 4.10.
Table 4.10.
Linear Regression Analyses for Main Effect of Flexible Workhours on Occupational Commitment
2005 (N = 823) 2015 (N = 749)
OC β Sig. △R2 β Sig. △R2
OC I (Job Change Difficulty) -.070 .069 .004 -.060 .131 .003 OC II (Job Change Possibility) -.006 .877 .000 -.023 .569 .000
OC III (Job Security) -.002 .964 .000 .025 .535 .001
Note. Independent Variable: Flexible Workhours. Dependent Variable: OC = Occupational Commitment.
H2-2. Flexible schedule is positively related to occupational commitment. In this study, the hypothesis 2-2 was used to test whether Flexible schedule was positively related to the three facets of OC. As the 2015 results, flexible schedule has a significant effect on OC III (β = .112, p < .05). The overall R2 of the 2015 results was .011 indicates that flexible schedule explained 1.1% of the variation in OC III. Thus, H2-2 was not supported by the 2005 data, but partially supported by 2015 in the present study, the results are presented in Table 4.11.
Table 4.11.
Linear Regression Analyses for Main Effect of Flexible Schedule on Occupational Commitment
2005 (N = 823) 2015 (N = 749)
OC β Sig. △R2 β Sig. △R2
OC I (Job Change Difficulty) .000 .994 .000 -.048 .230 .002 OC II (Job Change Possibility) .024 .541 .000 .029 .467 .001 OC III (Job Security) .069 .073 .004 .112* .005* .011 Note. Independent Variable: Flexible Schedule. Dependent Variable: OC = Occupational Commitment. *p < .05.
H2-3. Allowance for personal leave is positively related to occupational commitment. In this study, the hypothesis 2-3 was used to test whether allowance for personal leave was positively related to all three dimensions of OC. According to the 2005 and 2015 results, allowance for personal leave did not have significant effect on OC. Thus, H2-3 was not supported by the 2005 and 2015 data in the present study, the results are presented in Table 4.12.
Table 4.12.
Linear Regression Analyses for Main Effect of Allowance for Personal Leave on Occupational Commitment
2005 (N = 823) 2015 (N = 749)
OC β Sig. △R2 β Sig. △R2
OC I (Job Change Difficulty) .032 .386 .001 .065 .103 .004 OC II (Job Change Possibility) .027 .473 .001 .084 .036 .006
OC III (Job Security) .066 .074 .004 .028 480 .001
Note. Independent Variable: Allowance for Personal Leave. Dependent Variable: OC = Occupational Commitment.
Group Differences in Gender
The researcher also explored potential differences on FWA, JS and OC across gender using t test. The results of 2015 are presented in Table 4.13. For the research variables, males and females reported significant differences in means of flexible schedule (t = -2.090*), allowance for personal leave (t = 2.728**) and OC I (t = 2.061*). According to the means of flexible schedule, females (M = 2.09, SD = .78) reported higher scores than males (M = 1.97, SD = .80) that females accessed higher flexibility on their daily schedules. Based on the means of allowance for personal leave, males (M = 2.69, SD = 1.14) reported higher scores than females (M = 2.47, SD = 1.12) that males had less difficulty to take personal leave compared
to females. According to the means of OC I, males reported higher mean scores (M = 3.57, SD
= 1.28) than females’ (M = 3.38, SD = 1.28). Males considered they were not easy to find a job that is as good as current one, so they had higher commitment on their jobs than females.
Table 4.13.
The results of 2005 are presented in Table 4.14. For the research variables, the means of allowance for personal leave (t = 3.140**), JS II (t = 2.408*) and JS III (t = 2.755**) presented significant differences between two sexes. As the results showed that gender presented significant different mean scores of allowance for personal leave, males reported higher degree (M = 2.83, SD = 1.04) than females’ (M = 2.60, SD = 1.08) that that males had less difficulty to take personal leave compared to females. Gender also presented significant different mean scores of JS II and JS III. For JS II, males reported higher degree (M = 5.08, SD = 1.01) than females’ (M = 4.92, SD = .96) that men had higher satisfaction toward their main jobs than women. For JS III, males reported higher degree (M = 3.45, SD = .71) than females’ (M = 3.32,
SD = .67) that men had higher satisfaction toward their organizations than women.
Note. JS = Job Satisfaction; OC = Occupational Commitment. *p < .05, **p < .01.
The Moderating Effect of Gender
Hierarchical regression analysis was used to investigate the moderating effect of the moderator (gender) on the relationship between flexible work arrangements and job satisfaction, as well as the relationship between flexible work arrangements and occupational commitment, respectively. Three steps were conducted to investigate the moderating effect of gender on job satisfaction and occupational commitment. When testing one of the flexible work arrangements, the other two variables were placed at the first model as the control variables in this analysis, followed by the testing independent variables and gender in the second model, and finally the interaction terms were placed in the third model.
Table 4.15. shows the summary of the regression analyses of the moderating effect of gender on the relationships between FWA and JS. Results showed that whether in the results of 2005 or 2015, gender did not add significant incremental variance in each relationship
between FWA and JS. It indicated that males and females reported no different JS after accessing FWA. Therefore, the findings did not provide support for H3 that gender did not moderate the relationships between FWA and JS.
Table 4.15.
Summary of Hierarchical Regression Analyses for Moderating Effect of Gender on the Relationship between Flexible Work Arrangements and Job Satisfaction
2005 2015 Note. FWA = Flexible Work Arrangements; JS = Job Satisfaction.
Table 4.16. shows the summary of the regression analyses of the moderating effect of gender on the relationships between FWA and OC. As the results of 2005 showed that gender did not add any significant incremental variance to the relationships between allowance for personal leave and OC. However, as the results of 2015 presented that gender added significant incremental variances to the relationships between flexible workhours and OC I (β = .092, R2 = .008, p < .05), and OCII (β = .093, R2 = .009, p < .05), as well as the relationship between allowance for personal leave and OC I (β = .132, R2 = .017, p < .001).
Table 4.16.
Summary of Hierarchical Regression Analyses for Moderating Effect of Gender on the Relationship between Flexible Work Arrangements and Occupational Commitment
2005 2015
Note. FWA = Flexible Work Arrangements; OC = Occupational Commitment. *p < .05, ***p
< .001.
For the relationship between flexible workhours and OC I (job change difficulty), both males and females diminish their commitment after accessing flexible workhours, and males decreased more than females. It means that employees tend to reduce their occupational commitment and thought they could find another jobs as good as current one, when employers provided flexible workhours policy. Hence, flexible workhours produced a negative impact on job change difficulty for both men and women. Table 4.17. and Figure 4.1. outline the moderating effect of gender on the relationship between flexible workhours and OC I.
For the relationship between flexible workhours and OC II (job change possibility), males and females presented the opposite results. Even though these differences were impacted slightly, women increased their commitment toward their jobs, while males reduced after accessing flexible workhours. In other words, female employees were less likely to transfer to
another firms, while males increase the intention while employers offered flexible workhours policy. Hence, flexible workhours produced a positive impact on females’ commitment, but made a negative impact on males. Table 4.18. and Figure 4.2. outline the moderating effect of gender on the relationship between flexible workhours and OC II.
For the relationship between allowance for personal leave and OC I (job change difficulty), both males and females present positive results. Both men and women increased their commitment after receiving allowance for personal leave. Specifically, allowance for leave helped to improve job conditions that employees could not find better job than the current one
For the relationship between allowance for personal leave and OC I (job change difficulty), both males and females present positive results. Both men and women increased their commitment after receiving allowance for personal leave. Specifically, allowance for leave helped to improve job conditions that employees could not find better job than the current one