This chapter revealed the results after analyzing the data collected and some main findings related to work family conflict (WFC), family work conflict (FWC), workplace well-being, mindfulness, and family supportive organization perceptions (FSOP). In this session, descriptive statistics indicated the demographic characteristics about 306 participants. The tests of validity and reliability were proposed to make sure the quality of questionnaire. The correlation analysis showed the correlation among all the variables. Hierarchical regression analysis was presented to examine the hypotheses of the research. Confirmatory Factory Analysis was conducted to ensure the validity of each variable. Finally, the Common Method Variance (CMV) method was used to test the system errors in the measurements.
The hypotheses of this study were (1)Work family conflict is related to workplace well-being. (2) Family work conflict is related to workplace well-well-being. (3) Mindfulness will moderate the relationship between work family conflict and workplace well-being. (4) Mindfulness will moderate the relationship between family work conflict and workplace well-being. (5)Family supportive organization perceptions will moderate the relationship between work family conflict and workplace well-being. (6) Family supportive organization perceptions will moderate the relationship between family work conflict and workplace well-being.
Descriptive Statistics
The demographic information was conducted in nine items, gender, age, educational, marital status, child, seniority, position, income, and industrial in this study. The frequency and percentage of demographic information were presented in Table 4.1. The questionnaire was collected by on-line survey.
48
The total number of the participants was 306 employees from every walk of life. The respondents were consisted by 152 males (49.7%) and 154 females (50.3%). Most of them were 26 to 40 (34.6%) and 31 to 35 (23.9%) years old. The education levels were largely consisted by bachelor (41.8%) and master (39.5%) degree. There were 167 (54.6%) participants were single and the others (45.4%) were married. There were 137 (27.1%) participants with 1 to 2 kids and 157 (51.3%) with no kids. The most seniorities of participants were from 2 to 3 years (11.1%), 3 to 4 years (13.4%), 4 to 5 years (12.7%) and 10 years above (18.3%). 225 participants were non-management which accounted for 73.5% in the position. The average monthly income of most respondents were 27.8% from NT 25,001 to NT 40,000, 24.8% from NT 40,001 to NT 55,000 and 24.5% from NT 55,001 to NT 80,000. The major respondents came from three industries which are service industry (18.3%), high Tech manufacturing industry (23.9%) and the manufacturing industry (18.3%).
49
(continued) Table 4.1.
Demographic Information of the Sample (N = 306)
Variable Item Frequency Percentage (%)
1. Gender Female 154 50.3
50
Validity and Reliability
The Cronbach’s Alpha value can be seen as the reliability of variables. There were two versions of Cronbach’s alpha value in this study. Due to the analysis of factor loading, the researcher revised and deleted some low factor loading questions. Table 4.2. was the original one, and Table 4.3. was the revised one. The differences between Table 4.2. and Table 4.3. were
Table 4.1. (continued)
Variable Item Frequency Percentage (%)
7. Position Non-management 225 73.5
Management 81 26.5
Public Administration 32 10.5
Service 56 18.3
Financial and Insurance 19 6.2
Construction 11 3.6
Information Service 19 6.2
High Tech Manufacturing 73 23.9
Broadcasting 3 1
51
the values of workplace well-being, intrusion of work into private life and family supportive organization perceptions.
In Table 4.3., the scale showed the dependent variable, workplace well-being, had 30 items (deleted one low factor loading item) within four dimensions (work satisfaction, organizational respect for the employee, employer care, intrusion of work into private life). The overall Cronbach’s alpha of workplace well-being was 0.955. The questions of the independent variable (work family conflict, family work conflict) were from the same source. The Cronbach’s alpha of work family conflict was 0.912. The Cronbach’s alpha of family work conflict was 0.934. On the other hand, for the moderator parts, the scale of mindfulness was a 15-items under one dimension and the Cronbach’s alpha value was 0.944. Another moderator, family supportive organization perceptions with 10 items under one dimension, and the Cronbach’s alpha value was 0.860.
According to Tavakol and Dennick (2011), Cronbach’s Alpha value is acceptable while the value is over 0.70. In this study, most of the Cronbach’s Alpha value of variables were approximately over 0.90, which were indicated an acceptable result. Therefore, the scale applied in the study was no need to modify and the reliability were also acceptable.
52 Table 4.2.
Reliability Analysis of Each Variable - Original (N = 306)
Scale Dimension Number
of Items
Cronbach’s Alpha
Workplace Well-being 31 0.955
Work Satisfaction 10 0.942
Organizational Respect
Organization Perceptions 14 0.860
Table 4.3.
Reliability Analysis of Each Variable - Revised (N = 306)
Scale Dimension Number
of Items
Cronbach’s Alpha
Workplace Well-being 30 0.961
Work Satisfaction 10 0.942
Organizational Respect
Organization Perceptions 10 0.911
53
Common Method Variance (CMV)
In this research, the data collecting method was self-reported questionnaire. Namely, the respondent would answer all the questions from independent variables (work family conflict, family work conflict) to dependent variables (workplace well-being) even the moderators (mindfulness, family supportive organization perceptions). In this way, there might be possible with the CMV problems.
Podaskoff, Mackenzie, Lee, and Podsakoff (2003) contented that the research purpose and the name of variables should be displayed in the questionnaire. Moreover, the questionnaire also need to be designed in different Likert scale. For example, the work family conflict and family work conflict were 7-point Likert scale. Workplace well-being was a 5-point Likert scale. So as the moderator, mindfulness was a 6-point Likert scale, and family supportive organization perceptions (FSOP) was a 5-point Likert scale. By doing so to decrease the problems from CMV.
At the meantime, this study also adopted Harman’s one-factor analysis to test whether there was any CMV problem (Podsakoff & Organ, 1986). First, the researcher took all the questions of variable into the factor analysis process. Second, under the un-rotated situation, extract the first principle component of finding. If the extraction squared loading indicated the value under 50% which mean that there was no obvious CMV problem.
The finding of this study under un-rotated situation was 25.94% (the first main variable loading) with extraction 5 components. The results showed that there was no significant CMV problem.
Confirmatory Factor Analysis (CFA)
The study performed the confirmatory factor analysis (CFA) to examine the measurement validity. The measurement model fits of all instruments were conducted by AMOS. According
54
to the previous standard evaluation, the value of Chi-square divided by degree of freedom (ꭓ2/df ) should be under 3, which showed good model fit (Gefen, Straub, & Boudreau, 2000). Goodness of Fit Index (GFI), Adjusted Goodness of Fit Index (AGFI), Comparative Fit Index (CFI), Non-Normed Fit Index (TLI), Incremental Fit Index (IFI), Root Mean Square Error of Approximation (RMESA), and Standardized Root Mean Square Residual (SRMR) were the criteria to determine the overall model fit which indicated in Table 4.4.
Table 4.4.
The Criteria of Model Fit
Index Good Fit References
ꭓ2/df < 3 Gefen et al., 2000
GFI > 0.08 Byrne, 1994
AGFI > 0.90 Hu & Bentler, 1999
CFI > 0.90 Byrne, 1994
TLI > 0.90 Bagozzi & Yi, 1988
IFI > 0.90 Hu & Bentler, 1999
RMSEA < 0.08 McQuitty, 2004
SRMR < 0.08 Hu & Bentler, 1999
55
Convergent Validity
Composite reliability (CR) and average variance extracted (AVE) were implemented to evaluate the criteria of convergence validity. Composite reliability (CR) was the composition reliability of all measurement variables. Moreover, Fornell and Larcker (1981) and Bagozzi and Yi (1988) suggested that value higher than 0.60 was acceptable.
The value of composite reliability can be seen as similar to Cronbach’s alpha. It could be used to examine the reliability of overall variables (including discussing the model fit of dimension). Average variance extracted (AVE) is the variance measurement to evaluate the variable. Fornell and Larcker (1981) proposed that AVE is better to higher than 0.50. Higher AVE represented better convergent validity, the acceptable range is from 0.36 to 0.50.
56
The CFA Result of Work Family Conflict/ Family Work Conflict
The independent variable, work family conflict, family work conflict were overall 10 items. The confirmatory factor analysis (CFA) was performed in AMOS. The results showed in Table 4.5., 4.6., 4.7., 4.8., and CFA model showed in Figure 4.1. and 4.2.
Table 4.5.
Confirmatory Factor Analysis of Work Family Conflict (N = 306) Work Family Conflict done because of the demands my job puts on me.
0.89***
4 WF4 My job produces strain that makes it
difficult to fulfill family duties. 0.84***
5 WF5 Due to work-related duties, I have to make
changes to my plans for family activities. 0.77***
Note. ***p < .001
Table 4.6.
Model Fit Summary - Work Family Conflict (N = 306)
ꭓ2 df ꭓ2
/
df GFI AGFI RMSEA SRMR CFI TLI IFI27.335 5 5.467 .97 .90 .12 .03 .98 .96 .98
57 Table 4.7.
Confirmatory Factor Analysis of Family Work Conflict (N = 306) Family Work Conflict
Item
No. Code Item Factor
Loading CR AVE
6 FW1
The demands of my family or
spouse/partner interfere with work-related activities.
0.80***
0.93 0.74 7 FW2 I have to put off doing things at work
because of demands on my time at home. 0.86***
8 FW3
Things I want to do at work don't get done because of the demands of my family or spouse/partner.
0.86***
9 FW4
My home life interferes with my
responsibilities at work such as getting to work on time, accomplishing daily tasks, and working overtime.
0.90***
10 FW5 Family-related strain interferes with my
ability to perform job-related duties. 0.88***
Note. ***p < .001
Table 4.8.
Model Fit Summary - Family Work Conflict (N = 306)
ꭓ2 df ꭓ2
/
df GFI AGFI RMSEA SRMR CFI TLI IFI15.903 5 3.181 .98 .94 .09 .01 .99 .98 .99
58
Figure 4.1. Confirmatory factor analysis model of work family conflict
Figure 4.2. Confirmatory factor analysis model of family work conflict
59
The CFA Result of Workplace Well-being
The dependent variable, workplace well-being was an overall 31 items with four dimensions. The confirmatory factor analysis (CFA) was performed in AMOS. The results showed in Table 4.9., 4.10., and CFA model showed in Figure 4.3.
Although the AVE of intrusion of work into private life (AVE = .44) is lower than 0.5, with the comprehensive discussion and consideration, the researcher still kept this dimension in this variable. Due to the factor loading of each question in the dimension of intrusion of work into private was significant and over 0.5. The criteria still indicated that these problems were still reliable and valid.
60
(continued) Table 4.9.
Confirmatory Factor Analysis of Workplace Well-being (N = 306) - Revised Workplace Well-being give you a sense of direction and meaning?
0.82***
3 WS3_R Does your work bring a sense
of satisfaction? 0.89***
4 WS4_R Does your work increase your
sense of self-worth? 0.85***
5 WS5_R
Does your job allow you to re-craft your job to suit your needs? effective in your work on a day-to-day basis?
61
employer respects staff ? 0.83***
15 WO5_R How satisfied are you with
your work’s value system? 0.82***
16 WO6_R believe in the worth of the organization?
62
Note. ***p < .001; No.30 item (Do you feel that you can separate yourself easily from your work when you leave for the day?) was deleted due to the low factor loading.
Table 4.9. (continued)
Does your boss treat you as you would like to be
63 Table 4.10.
Model Fit Summary – Workplace Well-being (N = 306)
ꭓ2 df ꭓ2
/
df GFI AGFI RMSEA SRMR CFI TLI IFI1063.908 399 2.67 .81 .77 .07 .06 .91 .90 .91
Figure 4.3. Confirmatory factor analysis model of workplace well-being
64
The CFA Result of Mindfulness
The moderator, mindfulness was overall 15 items. The confirmatory factor analysis (CFA) was performed in AMOS. The results showed in Table 4.11., 4.12., and Figure 4.4.
65 Table 4.11.
Confirmatory Factor Analysis of Mindfulness (N = 306) Mindfulness carelessness, not paying attention, or thinking of something else.
0.72***
3 M3 I find it difficult to stay focused on what’s
happening in the present. 0.75***
4 M4 I tend to walk quickly to where I’m going
without paying attention along the way. 0.72***
5 M5
7 M7 It seems I’m “running on automatic” without
much awareness of what I’m doing. 0.84***
8 M8 I rush through activities without being really
attentive to them. 0.83***
9 M9
I get so focused on the goal I want to achieve that I lost touch with what I am doing right now to get there.
0.86***
10 M10 I do jobs or tasks automatically, without
being aware of what I’m doing. 0.86***
11 M11 I find myself listening to someone with one
ear, doing something else at the same time. 0.57***
12 M12 I drive places on “automatic pilot” and then
wonder why I went there. 0.81***
13 M13 I find myself preoccupied with the future or
the past. 0.53***
(continued)
66
Figure 4.4. Confirmatory factor analysis model of mindfulness Table 4.11. (continued)
Item
No. Code Item Factor
Loading CR AVE 14 M14 I find myself doing things without paying
attention. 0.82***
0.94 0.54 15 M15 I snack without being aware that I’m eating. 0.68***
Note. ***p < .001
Table 4.12.
Model Fit Summary - Mindfulness (N = 306)
ꭓ2 df ꭓ2
/
df GFI AGFI RMSEA SRMR CFI TLI IFI523.598 90 5.818 .80 .73 .13 .06 .87 .85 .87
67
The CFA Result of Family Supportive Organization Perceptions (FSOP)
The moderator, family supportive organization perceptions (FSOP) was overall 14 items with one reverse question. However, due to the low factor loading, there were four questions (Item 3.5.12.13.) were deleted. The confirmatory factor analysis (CFA) was performed in AMOS. The results showed in Table 4.13., 4.14., and Figure 4.5.
68 Table 4.13.
Confirmatory Factor Analysis of Family Supportive Organization Perceptions (FSOP) - Revised (N = 306)
Family Supportive Organization Perceptions (FSOP) Item
No. Code Item Factor
Loading CR AVE 1 FSOP1_R Work should be the primary priority
in a person’s life. 0.50***
0.91 0.51 2 FSOP2_R Long hours inside the office are the
way to achieving advancement. 0.67***
4 FSOP4_R It is considered taboo to talk about life
outside of work. 0.80***
Attending to personal needs, such as taking time off for sick children is frowned upon.
0.82***
8 FSOP8_R Employees should keep their personal
problems at home. 0.63***
9 FSOP9_R
The way to advance in this company is to keep non-work matters out of the workplace.
It is assumed that the most productive employees are those who put their work before their family life.
0.72***
14 FSOP14_R The ideal employee is the one who is
available 24 hours a day. 0.73***
Note. No.3.5.12.13. item was deleted due to the low factor loading; ***p < .001.
69
Figure 4.5. Confirmatory factor analysis model of family supportive organization perceptions (FSOP)
Table 4.14.
Model Fit Summary – Family Supportive Organization Perceptions (FSOP) (N = 306) ꭓ2 df ꭓ2
/
df GFI AGFI RMSEA SRMR CFI TLI IFI170.175 35 4.862 .90 .84 .11 .06 .92 .90 .91
70
Discriminant Validity
To have more comprehensive confirmation on the model, the researcher divided variables into different groups to discuss the measurement models. The models were conducted into 5-factor, 3-factor and 1-factor. Table 4.15. showed the 5-factor model (ꭓ2 = 6622.844,df = 2335, ꭓ2/df = 2.836, RMSEA = .08, CFI = .75, TLI = .75, IFI = .74) had an acceptable range. However, the criteria of CFI, TLI, and IFI in 5-factor model still could be improved. 1-factor model (ꭓ2
= 13301.026,df = 2345, ꭓ2/df = 5.67, RMSEA = .12, CFI = .36, TLI = .34, IFI = .36) was not in a good model fit. Therefore, the 5-factor model indicated the better fit compared to the 3-factor and 1-3-factor model.
Table 4.15.
Models Comparison of Fit Indices
5-Factor 3-Factor 1-Factor
ꭓ2 6622.844 7497.648 13301.026
df 2335 2342 2345
ꭓ2/df 2.836 3.201 5.67
RMSEA .08 .09 .12
CFI .75 .70 .36
TLI .75 .69 .34
IFI .74 .70 .36
Note. 5-factor model: Presented each measurement model; 3-factor model: WFC + FWC, FSOP + mindfulness, workplace well-being; 1-factor model: WFC + FWC + FSOP + mindfulness + workplace well-being
71
Pearson’s Correlation Analysis
The correlation analysis method was used to examine the correlation relationship among variables. The correlation relationship can provide two important statistic information. One is the direction of variables by using the symbol of (+) and (-); another is the association of variables by using the value of r. The mean, standard deviation, reliability and correlation were showed in Table 4.16.
According to Table 4.16., the dependent variable (workplace well-being) was significantly correlated to work family conflict (r = .341, p < .001); in the meantime, workplace well-being was significantly correlated to family work conflict (r = .339, p < .001). Therefore, the hypothesis 1 and 2 were both supported.
For the moderator, the mindfulness was significantly correlated to the work family conflict (r = .284, p < .001), family work conflict (r = .248, p < .001) and workplace well-being (r = .222, p < .001). And family supportive organization perceptions (FSOP) was significantly and negatively correlated to the work family conflict (r = -.144, p < .05), family work conflict (r = -.117, p < .05) and workplace well-being (r = -.160, p < .05) and mindfulness (r = -.661, p < .001).
72
Hierarchical Regression Analysis
The hierarchical regression was conducted to examine the relationship of moderating effect (Mindfulness, FSOP) on the work family conflict, family work conflict, and workplace well-being.
There were three models in the linear regression table. In the first step, the control variables (gender, marital, child) were controlled and examined in the first model. In the second step, the independent variables, work family conflict or family work conflict were put to check the effects of mindfulness or family supportive organization perceptions. For the last step, the interaction of the independent variable and moderator were put in the third model. In this way, total four regression analysis models were presented and reviewed all the hypothesizes in Table 4.17., 4.18., 4.19., 4.20.
According to Table 4.17., work family conflict (β = .358, p < .001) was positively and had a strong effect on workplace well-being. From Table 4.18., family work conflict (β = .375, p <
Table 4.16.
Means, Standard Deviations, Correlations and Reliability of Variables (N = 306)
Mean S.D. 1 2 3 4 5
1.WFC 4.63 1.42 (0.912)
2.FWC 4.51 1.47 .751*** (0.934)
3.WW 2.33 .71 .341*** .339*** (0.953)
4.MD 3.31 .99 .284*** .248*** .222*** (0.944)
5.FSOP 3.40 .85 -.144* -.117* -.160** .661*** (0.911) Note. *p < .05; **p < .01; ***p < .000.
WFC = Work Family Conflict; FWC = Family Work Conflict; WW = Workplace Well-being; MD = Mindfulness; FSOP = Family Supportive Organization Perceptions;
Numbers in parenthesis is Cronbach alpha value.
73
.001) was also positively and had a strong effect on workplace well-being. In this way, both hypothesis 1and hypothesis 2 were all accepted. Under the mindfulness as the moderator, Table 4.17. indicated that the interaction part is not significant in Model 3. So, hypothesis 3 was rejected. However, Table 4.18. indicated that mindfulness (β = 0.175, p < .01) as the moderator between the relationship family work conflict and workplace well-being was positively and significantly correlated. Therefore, hypothesis 4 was supported.
On the other hand, Table 4.19. and Table 4.20. presented FSOP as the moderator. There was no significant result when FSOP as the moderator between work family conflict and workplace well-being in Table 4.19. So, hypothesis 5 was rejected. Nevertheless, there was a significant result with the FSOP (β = -.163, p < .01) as the moderator in the relationship between family work conflict and workplace well-being (Table 4.20.). Finally, hypothesis 6 was supported.
74 Table 4.17.
Results of Hierarchical Regression Analysis in WFC - Mindfulness as the Moderator (N = 306)
Workplace Well-being
Variables Model 1 Model 2 Model3
Step 1
Gender 0.065 -0.022 -0.020
Marital 0.040 0.024 0.024
Child 0.114 0.105 0.094
Step 2
WFC 0.314*** 0.358***
Mindfulness 0.113* 0.094
Step 3
WFC × Mindfulness 0.107
R² 0.025 0.149 0.158
Adjusted R² 0.016 0.135 0.141
⊿R² 0.025 0.123 0.009
F 2.623 10.494*** 9.369***
⊿F 2.623 21.759*** 3.338
Note. *p < .05; ***p < .001; Two-tailed tests of significance.
75 Table 4.18.
Results of Hierarchical Regression Analysis in FWC - Mindfulness as the Moderator (N = 306)
Workplace Well-being
Variables Model 1 Model 2 Model3
Step 1
Gender 0.065 -0.007 0.001
Marital 0.040 0.021 0.002
Child 0.114 0.112 0.106
Step 2
FWC 0.311*** 0.375***
Mindfulness 0.124* 0.093
Step 3
FWC × Mindfulness 0.175**
R² 0.025 0.151 0.176
Adjusted R² 0.016 0.137 0.160
⊿R² 0.025 0.126 0.025
F 2.623 10.682*** 10.674***
⊿F 0.025 22.219*** 9.175**
Note. *p < .05; **p < .01; ***p < .001; Two-tailed tests of significance.
76 Table 4.19.
Results of Hierarchical Regression Analysis in WFC - FSOP as the Moderator (N = 306)
Workplace Well-being
Variables Model 1 Model 2 Model3
Step 1
Gender 0.065 -0.023 -0.024
Marital 0.040 0.025 0.044
Child 0.114 0.112 0.084
Step 2
WFC 0.332*** 0.375***
FSOP -0.095 -0.059
Step 3
WFC × FSOP -0.113
R² 0.025 0.146 0.156
Adjusted R² 0.016 0.132 0.139
⊿R² 0.025 0.121 0.010
F 2.623 10.281*** 9.206***
⊿F 2.604 0.348 7.759**
Note. **p < .01; ***p < .001; Two-tailed tests of significance.
77 Table 4.20.
Results of Hierarchical Regression Analysis in FWC - FSOP as the Moderator (N = 306)
Workplace Well-being
Variables Model 1 Model 2 Model3
Step 1
Gender 0.065 -0.007 0.001
Marital 0.040 0.022 0.025
Child 0.114 0.120 0.100
Step 2
FWC 0.330*** 0.378***
FSOP -0.104 -0.053
Step 3
FWC × FSOP -0.163**
R² 0.025 0.148 0.169
Adjusted R² 0.016 0.134 0.153
⊿R² 0.025 0.122 0.022
F 2.623 10.405*** 10.157***
⊿F 2.623 21.543*** 7.748**
Note. **p < .01; ***p < .001; Two-tailed tests of significance.
78
Figure 4.6. The moderating effects of mindfulness in the relationship between family work conflict and workplace well-being
From Figure 4.6., the mindfulness had a significant moderating effect on the family work conflict and workplace well-being. The mindfulness was divided into two groups, which mean the level of employee’s awareness toward mindfulness. The slope could explain the relationship between family work conflict and workplace well-being. The result indicated that there was a significant relationship between family work conflict and workplace well-being, which also supported the hypothesis 2. Moreover, the meaning of slope showed that even employee under high family work conflict situation, as they had high mindfulness, they still could experience high workplace well-being.
79
Figure 4.7. The moderating effects of family supportive organization perceptions (FSOP) in the relationship between family work conflict and workplace well-being
From Figure 4.7., the family supportive organization perceptions had a significant moderating effect on the family work conflict and workplace well-being. The family supportive organization perceptions were divided into two groups, which mean the level of employee’s perceptions toward the organization’s family support. The slope could explain the relationship between family work conflict and workplace well-being. The result indicated that there was a significant relationship between family work conflict and workplace well-being, which also supported the hypothesis 2. In addition, the meaning of slope showed that while employee under low family work conflict, with the help of high family supportive organization perceptions, they could experience high workplace well-being.
80
Discussion
The objective of the study was executed to inspect the relationship among work family conflict, family work conflict, and workplace well-being with the moderating effects on mindfulness and family supportive organization perceptions. All the hypotheses were executed and examined by correlation analysis and hierarchical regression analysis.
From the correlation analysis inspection, work family conflict had positive effect on workplace well-being. Besides, family work conflict also had positive effect on workplace well-being. In short, for the hypotheses 1 and 2 were both supported. In this study, WFC and FWC were both positively correlated with workplace well-being. Compared with other studies
From the correlation analysis inspection, work family conflict had positive effect on workplace well-being. Besides, family work conflict also had positive effect on workplace well-being. In short, for the hypotheses 1 and 2 were both supported. In this study, WFC and FWC were both positively correlated with workplace well-being. Compared with other studies