- Email: [email protected]
Non Standard Work Arrangements and Affective Commitment: The Mediating Role of Work-life Balance
Available online at www.sciencedirect.com
ScienceDirect Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
Evaluation of Learning for Performance Improvement International Conference Malaysia, 25 – 26 February, 2013
Non Standard Work Arrangements and Affective Commitment: the Mediating Role of Work-Life Balance Muhamad Khalil Omar* Universiti Teknologi MARA, Faculty of Business Management, Kuala Selangor 43200, Malaysia
Abstract This paper aims to determine the relationship of non-standard work arrangements such as non-standard work status, schedule, shift and hours towards work-life balance and affective commitment of services employees in Malaysia. Its secondary aim is to contribute to the literature by determining the mediating role of work-life balance in the relationship between non-standard work arrangements and affective commitment using analysis of Structural Equation Modelling. The results suggest that the preferences for standard or non-standard work arrangements are the elements which affect the employees’ satisfaction with work-life balance and affective commitment. The hypothesised model indicates the best fit for the mediating role of satisfaction with work-life balance.
© 2013The Published Elsevier Ltd. Selection and peer-review under responsibility of Universiti Malaysia © 2013 Authors. by Published by Elsevier Ltd. Malaysia Kelantan, Selection and peer-review under responsibility of Universiti Malaysia Kelantan, Malaysia Keywords: Non-standard work arrangements; work-life balance; affective commitment; non-standard employment
1. Introduction Current human resource practices have developed tremendously in last few decades due to the process of globalisation, technological advancement and social improvements. Work arrangements nowadays have been developed into two types, which are standard and non-standard. Standard work arrangement is traditional staffing practices involving the employment of permanent and full-time workers who are working in normal work schedules, shifts and hours (Walker, 2011). Whereas non-standard work arrangements are for modern staffing practices with the employment of non-permanent workers (i.e., parttime, contractual, temporary and etc) and flexible work schedules, shifts and hours (McNall, Masuda, & Nicklin, 2010). Polivka and Nardone (1989, p.11) defined non-standard employment as "...any job in which individual does not have an explicit or implicit contract for long-term employment or one in which *
Muhamad Khalil Omar. Tel.: +6-012-308-9147; fax: +6-035-544-4693. E-mail address: [email protected].
1877-0428 © 2013 The Authors. Published by Elsevier Ltd. Selection and peer-review under responsibility of Universiti Malaysia Kelantan, Malaysia doi:10.1016/j.sbspro.2013.12.392
Muhamad Khalil Omar / Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
the minimum hours worked can vary in a non-systematic way." Therefore, this study defines non-standard employment as any flexible work arrangements that do not include permanent full-time status and normal working schedule, shift and hours. 1.1 Non-standard work arrangements and affective commitment The studies of non-standard work arrangements have evolved since early 90’s as accordance with the advancement of human resource practices. Non-standard workers are acknowledged to differ demographically from standard employees, but the extent to which their work attitudes differ is less clear (Holtom, Lee, & Tidd, 2002). Hence the early studies of non-standard work arrangements were mostly concentrated in identifying the effect of non-standard employments towards work-related attitudes such as job satisfaction, commitment and turnover intention, and comparison with standard employees (Thorsteinson, 2003). However, past research findings comparing the job attitudes of standard and nonstandard workers have been conflicting and inconclusive (Conelly, Gallagher, & Giley, 2007). In terms of commitment levels of part-time and full-time employees, inconsistent findings have emerged as well. Studies have found part-time workers to be more affectively committed than full-timers (e.g. Martin & Sinclair, 2007). However there were studies that have revealed that part-timers to be less committed (e.g., Han, Moon, & Yun, 2009). Conversely, there were also studies that revealed part-timers as equally committed to their jobs as compared to full-time workers (e.g., Hill, Martinson, Ferris, & Baker, 2004). Studies comparing commitment levels between permanent and temporary workers also illustrated contradictory outcomes. Van Dyne and Ang (1998) found that temporary workers have more positive views of their psychological contracts and high affective commitment, as they view the flexibility of contingent work and their consequent ability to balance a professional career and their other life interests as important inducements by their organizations. In contrary, Coyle-Shapiro and Morrow (2006) concurred that permanent workers are more committed than temporary workers. In addition, Pearce (1993) did not find any significant difference in terms of commitment level between permanent and temporary or contract employees. To deal with all the discrepancies in prior research of non-standard work arrangements and its relationship with work-related attitudes, a new concept of work-status congruence based on discrepancy theory was used in this study. This new construct was developed by Holtom et al. (2002), and in their study, work status congruence is defined as the degree to which employers match employee preference for standard or non-standard work arrangements (i.e. work status, schedule, shift, and hours). Their studies had indicated that work status congruence was positively related with affective commitment and job satisfaction. Carr, Gregory and Harris (2010) validated Holtom et al.’s (2002) study and there were significant relationship between work status congruence and affective commitment as well as organisational citizenship behaviour. The work status congruence is theorised as a unifying concept to the inconsistent empirical findings regarding the attitudes of employees in standard and non-standard work arrangements. Besides, the results of prior studies also suggested that work status congruence is broader than a simple match between desired and actual staffing arrangements (e.g. full-time or part-time) and should include congruent preferences for scheduling arrangements (i.e. work schedule, shift and hours). However, in Holtom et al.’s (2002) study, work-status congruence using its new comprehensive measure was operationalised as full-time and part-time only without inclusion of temporary and permanent dimensions and its effect towards the work-life balance study was untested. The construct was also not being assessed as in hypothesized framework using structural equation modelling. Hence, this study beside to cope with the limitation of tested outcome variables, sample and analysis technique of work status congruence, it will further extend it towards work-life balance studies which has not been empirically examined before.
5
6
Muhamad Khalil Omar / Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
1.2 Non-standard work arrangements and work-life balance The articulation of work and life, cast as work-life balance, has become a key feature of much current government policymakers, practitioner and academic debate (Hyman & Summers, 2004). However, Moore (2007) explains that defining the "balance" in work/life is as unique to each individual as individuals are themselves. Recent changes in the nature of careers elevate a stronger concern for balancing work and non-work lives and flexible working options, which also known as non-standard work arrangements and are offered as work-life balance policies or initiatives allowing employers to appear employee-friendly whilst meeting business needs (Fleetwood, 2007). As concern for balancing work and non-work roles grows, work schedule flexibility, or the ease with which employees can change their work hours, may be a work characteristic that is increasingly favoured by employees (Jang, Park, & Zippay, 2011). Therefore, the use of several life-friendly policies and practices (e.g., flexible work schedules or nonstandard work arrangements) are likely to reduce work-family conflicts and personal stress as well as enhancing the work attitudes of employees (Sturges & Guest, 2004). Although the use of organisational work-life program has been shown to reduce work-family conflict (Aryee, Srivinas, & Tan, 2005), the studies of the effects of work-life balance by using flexible work options have been incoherent in their results and consequences. Higgins, Duxbury, and Johnson (2000) for instance, highlighted that nonstandard employment such as part-time offers the best of both worlds since it enables employees to pursue their career interests while still affording time to be with their families. On the contrary, it has been argued that the low fringe benefits, routine tasks, and limited career advancement opportunities that characterize most part-time jobs, make it more difficult for individuals to balance family demands (Hyman & Summers, 2004). Furthermore, Wayne, Randel, and Stevens (2006) also found that family/life-friendly benefits were not related with work-family enrichment. Additionally, Moore’s (2007) two-year-long ethnographic and in-depth interview study at an Anglo-German automobile factory did not found that flexible work arrangements were necessarily good for work-life balance. Instead it contributed to poor balance. Moore questioned whether flexible working practices have a positive or negative influence on work-life balance depends on the circumstances of the individual. Whereas, Baral and Bhargava (2010) discovered that there were no associations between the 22 items of work-life benefits and policies and work-family enrichment. Hence the utilisation of work status congruence in this study as a unifying concept of non-standard work arrangements (due to inclusion congruent preferences for work status, schedule, shift and hours) is expected to overcome the inconsistencies of the effects of non-standard employment towards employees’ work-life balance. Consequently, past studies have shown that work-life balance is positively related employees’ affective commitment (Sturges & Guest, 2004). Based on existing literatures, the following hypotheses are derived to identify each of the variable’s relationship in a framework; Hypotheses 1: There is a relationship between work status congruence and satisfaction with work-life balance; Hypotheses 2: There is a relationship between work status congruence and affective commitment; Hypotheses 3: There is a relationship between satisfaction with work-life balance and affective commitment; Hypotheses 4: The relationship between work status congruence and affective commitment is mediated by satisfaction with work-life balance. 2. Methodology This research was conducted in quantitative manner with concentration on survey method to enable it to be more conclusive and exclusive since the samples involved all dimensions of standard and nonstandard employment status including full-time and part-time as well as permanent and contractual or temporary basis. The method of sampling for this research was convenience sampling because of its procedures involve collecting information from members of the population who are conveniently available to provide pertinent information related to the study. The instrument for this research was a standardized questionnaire, as it provides an opportunity for the respondents to reply in full and honest
Muhamad Khalil Omar / Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
answers. The distribution of questionnaires was conducted in various organizations which represented various kinds of services and retail industries such as hotels, restaurants, hypermarkets, banking and hospitals. Additionally, data was collected from various ranks of employees in the organization ranging from non-clerical to lower managements. This is important so that the findings will be based on mixture of opinion regardless of employment status and to minimize bias of differences in task performed as these were among the limitations from the previous studies. In addition, the distribution of these questionnaires was done in various organizations located in Klang Valley area of Malaysia (covering territory of Kuala Lumpur and Selangor) since these areas are the most developed area in this country. Majority of companies operating under services sector has been setting up their main premises and headquarters here and employing most numbers of workers in this area. The collected data was analysed using Structural Equation Modelling (SEM), an analytic technique that provides an overall test of model fit and an assessment of model parameters (Byrne, 2010). Thus data could be processed comprehensively in order for the findings to be accurately presented in this study. By using SEM, various kinds of analysis methods was used to completely exploit the gathered data into a reliable results and workable solution as well as in order to ensure validation of measurements and confirmation of causality framework as hypothesized in order to achieve approachable conclusion and recommendations. The following measures were used in this study were adapted from various research, as this study will further validate the adapted measure as well as provides valuable contributions in extending prior works of studies in non-standard employment and work-life balance; work status congruence (Holtom et al., 2002). This 7-point scale was used to assess an employee’s congruent preferences for his/her work status, schedule, shift and hours. Satisfaction with work–life balance (Valcour, 2007). This 7-point scale was used to assess an employee’s satisfaction with his/her work-life balance. This scale was once called as satisfaction with work-family balance by Valcour (2007), but renamed as work-life in this study. This is because as an adapted scale, the instrument was adjusted to include “personal life” rather than using the word family only to make items equally relevant to respondents with and without family and to accommodate other non-work concerns of employees. Affective commitment was assessed with six-item measure developed by Meyer, Allen, and Smith (1993). Responses were based on a Likert scale (1 = strongly disagree, 7 = strongly agree). The alpha reliability coefficient was .85. 3. Results and Discussion Anderson and Gerbing’s (1988) “two-step approach” to structural equation modeling is used as a guideline for the data analysis. According to Anderson and Gerbing (1988), the first step requires development of a measurement model. The measurement model in the SEM model deals with the latent variables and their indicators. A pure measurement model is a confirmatory factor analysis (CFA) model in which there is unmeasured covariance between each possible pair of latent variables. The measurement model is evaluated like any other SEM model, using goodness of fit measures. Secondly, full structural model was executed, so that all hypothesised relationship between all latent variables and its observed variables can be tested, hence ascertaining the overall fit of the causal model towards the sample data. The AMOS program version 18.0 with maximum likelihood estimation was used to estimate the confirmatory and structural equation models in this study. Most of the respondents were in the age of between 20 to 29 years old (51%). Female dominated male respondents by 76% and most of the respondents were from the Malay ethnicity (60.7%), singles (53%), having secondary education or holding diploma/certificate (88%), working low levels (81%) and mostly full-time (76.6%). 3.1 Confirmatory factor analysis (CFA) A CFA was carried out for each of the three latent variables or construct of this study namely work status congruence, satisfaction with work life balance and affective commitment due to the reason for identifying only significant standardised parameter estimates as validity coefficient. Hence using this
7
8
Muhamad Khalil Omar / Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
analysis, the best indicator for a particular construct shall be determined. Next, full measurement model was developed to identify the overall fir of all variables under study; and at the same time identifying the discriminant validity for each construct by comparing it with other alternative models. After performing CFA for these three constructs, the fit indices generally indicate a not so well fitting model since the normed chi-square (5.464) is significant and above the recommended range. However other fit indices are as follows: RMSEA (.071), NFI (.883), TLI (.889), and CFI (.902). An examination of the AMOS output, including standardized factor weights and modification index, identified an error covariance between a few items i.e. e2 with e5, e13 with e14, e8 with e 9 and e15 with e18. This error covariance is justifiable as the items reflected a high level of similarity. The fit indices after model revised generally reflect a high level of fit. The absolute fit index of RMSEA (.054) provides evidence that the model fits the data well. The incremental or comparative fit indices also indicate a good fit with the TLI (.935), CFI (.944), and NFI (.925) being above the recommended minimum value of .90. The normed chi- square (CMIN/DF) of 3.606 also near the recommended range of maximum three. 3.2 Full measurement model In this study, work status congruence, satisfaction with work-life balance and affective commitment were treated as a 3 Factor model considering the differences involved. Based on AMOS results, this measurement model was adequate and has fit the data very well since all the indices were above minimum requirement for best fit model, i.e TLI (.935), CFI (.944), NFI (.925), RMSEA (.54) and normed chi- square (CMIN/DF) of 3.606. However, before the testing of the next step of hypothesised structural model is conducted, it was essential to identify construct validity by comparing this measurement model with other alternative models. This shall be gained by assessing the discriminant validity, which refers to the extent to which a certain construct is different from other constructs. Hence, in this full measurement model, all the three constructs (i.e. 3 Factor Model) need to be tested for discriminant validity, so that it can verify that the scales developed to measure different constructs, are indeed measuring different constructs (Byrne, 2010). Consequently, the discriminant validity of each construct shall be ascertained using model comparison with 2 Factor Model (if there were 2 groups of construct which were discriminant, i.e work status congruence and satisfaction with work-life balance were in 1 group of construct and affective commitment was in the other) and 1 Factor Model (if none of the construct were discriminant i.e work status congruence, satisfaction with work-life balance and affective commitment were the same). Based on AMOS results, this measurement model was adequate and has fit the data very well since all the indices were above minimum requirement for best fit model, i.e TLI (.935), CFI (.944), NFI (.925), RMSEA (.54) and normed chi-square (CMIN/DF) of 3.606. However, before the testing of the next step of hypothesised structural model is conducted, it was essential to identify construct validity of this full measurement model which shall be achieved by comparing this measurement model with other alternative models. This shall be gained by assessing the discriminant validity which refers to the extent to which a certain construct is different from other constructs. Hence in this full measurement model, all 3 constructs need to be tested for discriminant validity so that it can verify that the scales developed to measure different constructs, are indeed measuring different constructs (Byrne, 2010). Hence, this discriminant validity of each construct shall be ascertained using model comparison with 2 Factor Model (if there were 2 groups of construct which were discriminant i.e work status congruence and satisfaction with work-life balance were in 1 group of construct and affective commitment was in the other) and 1 Factor Model (if none of the construct were discriminant i.e work status congruence, satisfaction with work-life balance and affective commitment were the same).
9
Muhamad Khalil Omar / Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
Table 1. Model fit summary (model comparison for measurement) Model Model 1 (1 Factor) Model 2 (2 Factor) Model 3 (3 Factor) Differences (Model 3-1) Differences (Model 3-2) Note: * p≤0.001
ᵡ2 739.698 673.424 587.713
df 166 164 163 3 1
Δᵡ2
151.985* 85.711*
ᵡ2/df 4.456 4.106 3.606
NFI .905 .914 .925
CFI .925 .933 .944
TLI .914 .923 .935
RMSEA .062 .059 .054
Based on Table 1, it was found that the 3-Factor Model which is the full measurement model under study was the best model that mostly fitted the samples data very well. This model was having the smallest χ², CMIN/DF and RMSEA. In addition, the hypothesized measurement model was the best fit since all of its goodness-of-fit were more than the required value and the highest among other model for fine-fitting model (i.e. NFI, CFI and TLI of > .90). Hence, this model was more superior as compared to other models. Furthermore, in assessing the extent to which a model exhibited the best fit, a univariate approach was used to determine if the difference in fit between the three models were statistically significant. As such, the differences in χ² (Δχ²) values between the three models was examined with the presumption that all models were nested and a significant Δχ² indicating substantial improvement in model fit. Comparison of 3-Factor Model with other models have yielded a differences in χ² of 151.985 if to compare with 1-Factor Model and differences of 85.711 if to compare with 2-Factor Model; while all the differences in χ² (Δχ²) were significant since p-value was < .000. Hence it was certain that all of these constructs in the full measurement model under study were having discriminant validity, thus it was confirmed that the scales that were developed to measure different constructs, were indeed measuring different constructs. 3.3 Full structural model Since the full measurement model was successfully conducted for each of constructs under this study, and it was satisfied that the measurement model was valid and well fitted to the data, thus it was best to proceed to full structural model to test the hypothesized research model between latent variables with full SEM. By executing the full structural model, then the hypothesised relationship between latent variables can be tested hence ascertaining the overall fit of the proposed model towards the samples data. In addition, once the full structural model was satisfied, the mediation effect shall be further testified against the alternatives. Based on the results of full measurement model and the proposed research framework as explained in the Methodology section, the full structural model was developed as per Figure 1. Based on Figure 1, certain constructs were modified during CFA and measurement model, and the relationship of each of the variables and its indicators were examined to test the hypotheses as well as to identify whether the proposed model fit the data well. As shown in Figure 1, satisfaction with work-life balance was hypothesised as partial mediating variables between work status congruence and Affective Commitment. Hence, this proposed model need to be verified against other alternative models i.e. Full Mediation Model (no direct relationship between work status congruence and Affective Commitment) and Non-Mediation Model (no direct relationship between affective commitment and Job Satisfaction). By accomplishing this model comparison verification, the best model shall be ascertained hence testifying the proposed partial mediation model and at the same time proving the whole model fit towards data sample. Based on results in estimates, it was found that all variables and their indicators were significantly related in a positive way. Hence, it is confirmed that there is a relationship between work status congruence and satisfaction with work-life balance hence supporting Hypotheses 1 with a significant standardised regression weight of .509. Similarly for Hypotheses 2, it is verified that there is a relationship between work status congruence and affective commitment with a significant standardised
10
Muhamad Khalil Omar / Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
regression weight of .253. As for Hypotheses 3, it was testified that there is a relationship between satisfaction with work-life balance and affective commitment with a significant standardised regression weight of .615.
e21 WLB7
1 1
Satisfaction with Work-Life Balance e1
e2 e3 e4 e5 e6
e7
1
1 1
1 1 1
1
WSC1
WSC4
1
WLB5
1
WLB4
1
WLB2
WSC3 Work Status Congruence
e14
WLB6
WLB3
WSC2
1
WLB1
e13 e12
e11
1
e10
1
e9
1
e8
WSC5 WSC6 WSC7
1
AC6 AC5
Affective Commitment
AC4 AC3
e22
AC2 AC1
1
e20
1 e19 1
e18
1 e17 1
1
e16 e15
Figure 1: Full structural model
Based on Table 2, it was found that the proposed Partial Mediation Model which is the full structural model under study was the best model that mostly fitted the samples data very well. This model was having the smallest χ², CMIN/DF and RMSEA. In addition, the hypothesised partial mediation model was the best fit since all of its goodness-of-fit were more than the required value and the highest among other model for fine-fitting model (i.e. NFI, CFI and TLI of > .90). Hence this model was more superior as compared to other models thus supporting Hypotheses 4 that the relationship between work status congruence and affective commitment is mediated by satisfaction with work-life balance. Furthermore, in assessing the extent to which a model exhibited the best fit, a univariate approach was used to determine if the difference in fit between the three models were statistically significant. As such, the differences in χ² (Δχ²) values between the three models was examined and in comparison with 2 other models have yielded a differences in χ² of 80.915 if to compare with Full Mediation Model and differences of 11.448 if
11
Muhamad Khalil Omar / Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
to compare with Non Mediation Model; while all the differences in χ² (Δχ²) were significant since p-value was .001. Table 2. Model fit summary (model comparison for mediation) Model Model 1 (Partial Mediation) Model 2 (Full Mediation) Model 3 (Non mediation) Differences (Model 3-1) Differences (Model 3-2) Note: * p≤0.001
ᵡ2 587.713 668.628 599.161
Df 163 164 164 1 1
Δᵡ2
80.915* 11.448*
ᵡ2/df 3.606 4.077 3.653
NFI .925 .914 .923
CFI .944 .934 .943
TLI .935 .934 .943
RMSEA .054 .059 .055
4. Conclusions This research has made its contributions with regard to the body of knowledge especially in extending the concept of work status congruence in work-life balance studies. Moreover, operationalisations of work status congruence and satisfaction with work-life balance were enhanced by inclusion of other nonwork activities and expanded on the subject of non-standard and temporary workers. Hence, this research has complemented prior studies of non-standard employment and work-life balance especially in substantiating positive impact of both domains towards affective commitment. Furthermore, effect of satisfaction with work-life balance as a mediator in work status congruence and affective commitment relationship has been established thus supporting previous scholars’ call for such confirmation. References Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: A review and recommend two-step approach. Psychological Bulletin, 103(3), 411-423. Aryee, S., Srinivas, E. S., & Tan, H. H. (2005). Rhythms of life: Antecedents and outcomes of work– family balance in employed parents. Journal of Applied Psychology, 90, 132–146. Baral, R., & Bhargava, S. (2010). Work-family enrichment as a mediator between organizational interventions for work-life balance and job outcomes. Journal of Managerial Psychology, 25(3), 274300. Byrne, B. M. (2010). Structural equation modeling with AMOS: Basic concepts, applications, and programming (2nd ed.). NY: Routledge Academic. Carr, J. C., Gregory, B. T., & Harris, S. G. (2010). Work status congruence’s relation to employee attitudes and behaviors: The moderating role of procedural justice. Journal of Business Psychology, 25, 583–592. DOI 10.1007/s10869-009-9151-z Connelly, C. E., Gallagher, D. G., & Gilley, K. M. (2007). Organizational and client commitment among contracted employees: A replication and extension with temporary workers. Journal of Vocational Behavior, 69, 326–335. Coyle-Shapiro, J. A-M., & Morrow, P. C. (2006). Organisational and client commitment among contracted employees. Journal of Vocational Behaviour, 68, 416-431. Fleetwood, S. (2007). Why work–life balance now? International Journal of Human Resource Management, 18(3), 387–400. Han, S.-S., Moon, S. J., & Yun, E. K. (2009). Empowerment, job satisfaction, and organizational commitment: Comparison of permanent and temporary nurses in Korea. Applied Nursing Research, 22, 15–20. Higgins, C., Duxbury, L., & Johnson, K. L. (2000). Part-time work for women: Does it really help balance work and family? Human Resource Management, 39(1), 17–32.
12
Muhamad Khalil Omar / Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
Hill, E. J., Ferris, M., & Martinson, V. (2003). Does it matter where you work? a comparison of how three work venues (traditional office, virtual office, and home office) influence aspects of work and personal/family life. Journal of Vocational Behavior, 63(2), 220–241. Holtom, B. C., Lee, T. W., & Tidd, S. T. (2002). The relationship between work status congruence and work-related attitudes and behaviors. Journal of Applied Psychology, 87(5), 903–915. Hyman, J., & Summers, J. (2004). Lacking balance? Work-life employment practices in the modern economy. Personnel Review, 33(4), 418-429. Jang, S. J., Park, R., & Zippay, A. (2011). The interaction effects of scheduling control and work–life balance programs on job satisfaction and mental health. International Journal Social Welfare, 20, 135–143. DOI: 10.1111/j.1468-2397.2010.00739.x Martin, J. E., & Sinclair, R. R. (2007). A typology of the part-time workforce: Differences on job attitudes and turnover. Journal of Occupational and Organizational Psychology, 80(2), 301–319. McNall, L. A., Masuda, A. D., & Nicklin, J. M. (2010). Flexible work arrangements, job satisfaction, and turnover intentions: The mediating role of work-to-family enrichment. The Journal of Psychology, 144(1), 61–81. Meyer, J. P., Allen, N. J., & Smith, C. A. (1993). Commitment to organizations and occupations: extension and test of a three-component conceptualization. Journal of Applied Psychology, 78, 538– 551. Moore, F. (2007). Work-life balance: contrasting managers and workers in an MNC. Employee Relations, 29(4), 385-399. Pearce, J. L. (1993). Toward an organizational behavior of contract laborers: Their psychological involvement and effects on employee coworkers. Academy of Management Journal, 36, 1082–1096. Polivka, A. E., & Nardone, T. (1989). The definition of contingent work. Monthly Labor Review, 112, 916. Sturges, J., & Guest, D. (2004). Working to live or living to work? Work/life balance early in the career. Human Resource Management Journal, 14(4), 5-20. Thorsteinson, T. J. (2003). Job attitudes of part-time versus full-time workers: a meta analytic review. Journal of Occupational and Organizational Psychology, 76, 151-77. Valcour, M. (2007). Work-based resources as moderators of the relationship between work hours and satisfaction with work–family balance. Journal of Applied Psychology, 92(6), 1512–1523. Van Dyne, L., & Ang, S. (1998). Organizational citizenship behavior of contingent workers in Singapore. Academy of Management Journal, 41, 692–703. Walker, B. (2011). How does non-standard employment affect workers? A consideration of the evidence. New Zealand Journal of Employment Relations, 36(3), 14-29. Wayne, J. H., Randel, A. E., & Stevens, J. (2006). The role of identity and work–family support in work– family enrichment and its work-related consequences. Journal of Vocational Behaviour, 69(3), 445461.
ScienceDirect Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
Evaluation of Learning for Performance Improvement International Conference Malaysia, 25 – 26 February, 2013
Non Standard Work Arrangements and Affective Commitment: the Mediating Role of Work-Life Balance Muhamad Khalil Omar* Universiti Teknologi MARA, Faculty of Business Management, Kuala Selangor 43200, Malaysia
Abstract This paper aims to determine the relationship of non-standard work arrangements such as non-standard work status, schedule, shift and hours towards work-life balance and affective commitment of services employees in Malaysia. Its secondary aim is to contribute to the literature by determining the mediating role of work-life balance in the relationship between non-standard work arrangements and affective commitment using analysis of Structural Equation Modelling. The results suggest that the preferences for standard or non-standard work arrangements are the elements which affect the employees’ satisfaction with work-life balance and affective commitment. The hypothesised model indicates the best fit for the mediating role of satisfaction with work-life balance.
© 2013The Published Elsevier Ltd. Selection and peer-review under responsibility of Universiti Malaysia © 2013 Authors. by Published by Elsevier Ltd. Malaysia Kelantan, Selection and peer-review under responsibility of Universiti Malaysia Kelantan, Malaysia Keywords: Non-standard work arrangements; work-life balance; affective commitment; non-standard employment
1. Introduction Current human resource practices have developed tremendously in last few decades due to the process of globalisation, technological advancement and social improvements. Work arrangements nowadays have been developed into two types, which are standard and non-standard. Standard work arrangement is traditional staffing practices involving the employment of permanent and full-time workers who are working in normal work schedules, shifts and hours (Walker, 2011). Whereas non-standard work arrangements are for modern staffing practices with the employment of non-permanent workers (i.e., parttime, contractual, temporary and etc) and flexible work schedules, shifts and hours (McNall, Masuda, & Nicklin, 2010). Polivka and Nardone (1989, p.11) defined non-standard employment as "...any job in which individual does not have an explicit or implicit contract for long-term employment or one in which *
Muhamad Khalil Omar. Tel.: +6-012-308-9147; fax: +6-035-544-4693. E-mail address: [email protected].
1877-0428 © 2013 The Authors. Published by Elsevier Ltd. Selection and peer-review under responsibility of Universiti Malaysia Kelantan, Malaysia doi:10.1016/j.sbspro.2013.12.392
Muhamad Khalil Omar / Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
the minimum hours worked can vary in a non-systematic way." Therefore, this study defines non-standard employment as any flexible work arrangements that do not include permanent full-time status and normal working schedule, shift and hours. 1.1 Non-standard work arrangements and affective commitment The studies of non-standard work arrangements have evolved since early 90’s as accordance with the advancement of human resource practices. Non-standard workers are acknowledged to differ demographically from standard employees, but the extent to which their work attitudes differ is less clear (Holtom, Lee, & Tidd, 2002). Hence the early studies of non-standard work arrangements were mostly concentrated in identifying the effect of non-standard employments towards work-related attitudes such as job satisfaction, commitment and turnover intention, and comparison with standard employees (Thorsteinson, 2003). However, past research findings comparing the job attitudes of standard and nonstandard workers have been conflicting and inconclusive (Conelly, Gallagher, & Giley, 2007). In terms of commitment levels of part-time and full-time employees, inconsistent findings have emerged as well. Studies have found part-time workers to be more affectively committed than full-timers (e.g. Martin & Sinclair, 2007). However there were studies that have revealed that part-timers to be less committed (e.g., Han, Moon, & Yun, 2009). Conversely, there were also studies that revealed part-timers as equally committed to their jobs as compared to full-time workers (e.g., Hill, Martinson, Ferris, & Baker, 2004). Studies comparing commitment levels between permanent and temporary workers also illustrated contradictory outcomes. Van Dyne and Ang (1998) found that temporary workers have more positive views of their psychological contracts and high affective commitment, as they view the flexibility of contingent work and their consequent ability to balance a professional career and their other life interests as important inducements by their organizations. In contrary, Coyle-Shapiro and Morrow (2006) concurred that permanent workers are more committed than temporary workers. In addition, Pearce (1993) did not find any significant difference in terms of commitment level between permanent and temporary or contract employees. To deal with all the discrepancies in prior research of non-standard work arrangements and its relationship with work-related attitudes, a new concept of work-status congruence based on discrepancy theory was used in this study. This new construct was developed by Holtom et al. (2002), and in their study, work status congruence is defined as the degree to which employers match employee preference for standard or non-standard work arrangements (i.e. work status, schedule, shift, and hours). Their studies had indicated that work status congruence was positively related with affective commitment and job satisfaction. Carr, Gregory and Harris (2010) validated Holtom et al.’s (2002) study and there were significant relationship between work status congruence and affective commitment as well as organisational citizenship behaviour. The work status congruence is theorised as a unifying concept to the inconsistent empirical findings regarding the attitudes of employees in standard and non-standard work arrangements. Besides, the results of prior studies also suggested that work status congruence is broader than a simple match between desired and actual staffing arrangements (e.g. full-time or part-time) and should include congruent preferences for scheduling arrangements (i.e. work schedule, shift and hours). However, in Holtom et al.’s (2002) study, work-status congruence using its new comprehensive measure was operationalised as full-time and part-time only without inclusion of temporary and permanent dimensions and its effect towards the work-life balance study was untested. The construct was also not being assessed as in hypothesized framework using structural equation modelling. Hence, this study beside to cope with the limitation of tested outcome variables, sample and analysis technique of work status congruence, it will further extend it towards work-life balance studies which has not been empirically examined before.
5
6
Muhamad Khalil Omar / Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
1.2 Non-standard work arrangements and work-life balance The articulation of work and life, cast as work-life balance, has become a key feature of much current government policymakers, practitioner and academic debate (Hyman & Summers, 2004). However, Moore (2007) explains that defining the "balance" in work/life is as unique to each individual as individuals are themselves. Recent changes in the nature of careers elevate a stronger concern for balancing work and non-work lives and flexible working options, which also known as non-standard work arrangements and are offered as work-life balance policies or initiatives allowing employers to appear employee-friendly whilst meeting business needs (Fleetwood, 2007). As concern for balancing work and non-work roles grows, work schedule flexibility, or the ease with which employees can change their work hours, may be a work characteristic that is increasingly favoured by employees (Jang, Park, & Zippay, 2011). Therefore, the use of several life-friendly policies and practices (e.g., flexible work schedules or nonstandard work arrangements) are likely to reduce work-family conflicts and personal stress as well as enhancing the work attitudes of employees (Sturges & Guest, 2004). Although the use of organisational work-life program has been shown to reduce work-family conflict (Aryee, Srivinas, & Tan, 2005), the studies of the effects of work-life balance by using flexible work options have been incoherent in their results and consequences. Higgins, Duxbury, and Johnson (2000) for instance, highlighted that nonstandard employment such as part-time offers the best of both worlds since it enables employees to pursue their career interests while still affording time to be with their families. On the contrary, it has been argued that the low fringe benefits, routine tasks, and limited career advancement opportunities that characterize most part-time jobs, make it more difficult for individuals to balance family demands (Hyman & Summers, 2004). Furthermore, Wayne, Randel, and Stevens (2006) also found that family/life-friendly benefits were not related with work-family enrichment. Additionally, Moore’s (2007) two-year-long ethnographic and in-depth interview study at an Anglo-German automobile factory did not found that flexible work arrangements were necessarily good for work-life balance. Instead it contributed to poor balance. Moore questioned whether flexible working practices have a positive or negative influence on work-life balance depends on the circumstances of the individual. Whereas, Baral and Bhargava (2010) discovered that there were no associations between the 22 items of work-life benefits and policies and work-family enrichment. Hence the utilisation of work status congruence in this study as a unifying concept of non-standard work arrangements (due to inclusion congruent preferences for work status, schedule, shift and hours) is expected to overcome the inconsistencies of the effects of non-standard employment towards employees’ work-life balance. Consequently, past studies have shown that work-life balance is positively related employees’ affective commitment (Sturges & Guest, 2004). Based on existing literatures, the following hypotheses are derived to identify each of the variable’s relationship in a framework; Hypotheses 1: There is a relationship between work status congruence and satisfaction with work-life balance; Hypotheses 2: There is a relationship between work status congruence and affective commitment; Hypotheses 3: There is a relationship between satisfaction with work-life balance and affective commitment; Hypotheses 4: The relationship between work status congruence and affective commitment is mediated by satisfaction with work-life balance. 2. Methodology This research was conducted in quantitative manner with concentration on survey method to enable it to be more conclusive and exclusive since the samples involved all dimensions of standard and nonstandard employment status including full-time and part-time as well as permanent and contractual or temporary basis. The method of sampling for this research was convenience sampling because of its procedures involve collecting information from members of the population who are conveniently available to provide pertinent information related to the study. The instrument for this research was a standardized questionnaire, as it provides an opportunity for the respondents to reply in full and honest
Muhamad Khalil Omar / Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
answers. The distribution of questionnaires was conducted in various organizations which represented various kinds of services and retail industries such as hotels, restaurants, hypermarkets, banking and hospitals. Additionally, data was collected from various ranks of employees in the organization ranging from non-clerical to lower managements. This is important so that the findings will be based on mixture of opinion regardless of employment status and to minimize bias of differences in task performed as these were among the limitations from the previous studies. In addition, the distribution of these questionnaires was done in various organizations located in Klang Valley area of Malaysia (covering territory of Kuala Lumpur and Selangor) since these areas are the most developed area in this country. Majority of companies operating under services sector has been setting up their main premises and headquarters here and employing most numbers of workers in this area. The collected data was analysed using Structural Equation Modelling (SEM), an analytic technique that provides an overall test of model fit and an assessment of model parameters (Byrne, 2010). Thus data could be processed comprehensively in order for the findings to be accurately presented in this study. By using SEM, various kinds of analysis methods was used to completely exploit the gathered data into a reliable results and workable solution as well as in order to ensure validation of measurements and confirmation of causality framework as hypothesized in order to achieve approachable conclusion and recommendations. The following measures were used in this study were adapted from various research, as this study will further validate the adapted measure as well as provides valuable contributions in extending prior works of studies in non-standard employment and work-life balance; work status congruence (Holtom et al., 2002). This 7-point scale was used to assess an employee’s congruent preferences for his/her work status, schedule, shift and hours. Satisfaction with work–life balance (Valcour, 2007). This 7-point scale was used to assess an employee’s satisfaction with his/her work-life balance. This scale was once called as satisfaction with work-family balance by Valcour (2007), but renamed as work-life in this study. This is because as an adapted scale, the instrument was adjusted to include “personal life” rather than using the word family only to make items equally relevant to respondents with and without family and to accommodate other non-work concerns of employees. Affective commitment was assessed with six-item measure developed by Meyer, Allen, and Smith (1993). Responses were based on a Likert scale (1 = strongly disagree, 7 = strongly agree). The alpha reliability coefficient was .85. 3. Results and Discussion Anderson and Gerbing’s (1988) “two-step approach” to structural equation modeling is used as a guideline for the data analysis. According to Anderson and Gerbing (1988), the first step requires development of a measurement model. The measurement model in the SEM model deals with the latent variables and their indicators. A pure measurement model is a confirmatory factor analysis (CFA) model in which there is unmeasured covariance between each possible pair of latent variables. The measurement model is evaluated like any other SEM model, using goodness of fit measures. Secondly, full structural model was executed, so that all hypothesised relationship between all latent variables and its observed variables can be tested, hence ascertaining the overall fit of the causal model towards the sample data. The AMOS program version 18.0 with maximum likelihood estimation was used to estimate the confirmatory and structural equation models in this study. Most of the respondents were in the age of between 20 to 29 years old (51%). Female dominated male respondents by 76% and most of the respondents were from the Malay ethnicity (60.7%), singles (53%), having secondary education or holding diploma/certificate (88%), working low levels (81%) and mostly full-time (76.6%). 3.1 Confirmatory factor analysis (CFA) A CFA was carried out for each of the three latent variables or construct of this study namely work status congruence, satisfaction with work life balance and affective commitment due to the reason for identifying only significant standardised parameter estimates as validity coefficient. Hence using this
7
8
Muhamad Khalil Omar / Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
analysis, the best indicator for a particular construct shall be determined. Next, full measurement model was developed to identify the overall fir of all variables under study; and at the same time identifying the discriminant validity for each construct by comparing it with other alternative models. After performing CFA for these three constructs, the fit indices generally indicate a not so well fitting model since the normed chi-square (5.464) is significant and above the recommended range. However other fit indices are as follows: RMSEA (.071), NFI (.883), TLI (.889), and CFI (.902). An examination of the AMOS output, including standardized factor weights and modification index, identified an error covariance between a few items i.e. e2 with e5, e13 with e14, e8 with e 9 and e15 with e18. This error covariance is justifiable as the items reflected a high level of similarity. The fit indices after model revised generally reflect a high level of fit. The absolute fit index of RMSEA (.054) provides evidence that the model fits the data well. The incremental or comparative fit indices also indicate a good fit with the TLI (.935), CFI (.944), and NFI (.925) being above the recommended minimum value of .90. The normed chi- square (CMIN/DF) of 3.606 also near the recommended range of maximum three. 3.2 Full measurement model In this study, work status congruence, satisfaction with work-life balance and affective commitment were treated as a 3 Factor model considering the differences involved. Based on AMOS results, this measurement model was adequate and has fit the data very well since all the indices were above minimum requirement for best fit model, i.e TLI (.935), CFI (.944), NFI (.925), RMSEA (.54) and normed chi- square (CMIN/DF) of 3.606. However, before the testing of the next step of hypothesised structural model is conducted, it was essential to identify construct validity by comparing this measurement model with other alternative models. This shall be gained by assessing the discriminant validity, which refers to the extent to which a certain construct is different from other constructs. Hence, in this full measurement model, all the three constructs (i.e. 3 Factor Model) need to be tested for discriminant validity, so that it can verify that the scales developed to measure different constructs, are indeed measuring different constructs (Byrne, 2010). Consequently, the discriminant validity of each construct shall be ascertained using model comparison with 2 Factor Model (if there were 2 groups of construct which were discriminant, i.e work status congruence and satisfaction with work-life balance were in 1 group of construct and affective commitment was in the other) and 1 Factor Model (if none of the construct were discriminant i.e work status congruence, satisfaction with work-life balance and affective commitment were the same). Based on AMOS results, this measurement model was adequate and has fit the data very well since all the indices were above minimum requirement for best fit model, i.e TLI (.935), CFI (.944), NFI (.925), RMSEA (.54) and normed chi-square (CMIN/DF) of 3.606. However, before the testing of the next step of hypothesised structural model is conducted, it was essential to identify construct validity of this full measurement model which shall be achieved by comparing this measurement model with other alternative models. This shall be gained by assessing the discriminant validity which refers to the extent to which a certain construct is different from other constructs. Hence in this full measurement model, all 3 constructs need to be tested for discriminant validity so that it can verify that the scales developed to measure different constructs, are indeed measuring different constructs (Byrne, 2010). Hence, this discriminant validity of each construct shall be ascertained using model comparison with 2 Factor Model (if there were 2 groups of construct which were discriminant i.e work status congruence and satisfaction with work-life balance were in 1 group of construct and affective commitment was in the other) and 1 Factor Model (if none of the construct were discriminant i.e work status congruence, satisfaction with work-life balance and affective commitment were the same).
9
Muhamad Khalil Omar / Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
Table 1. Model fit summary (model comparison for measurement) Model Model 1 (1 Factor) Model 2 (2 Factor) Model 3 (3 Factor) Differences (Model 3-1) Differences (Model 3-2) Note: * p≤0.001
ᵡ2 739.698 673.424 587.713
df 166 164 163 3 1
Δᵡ2
151.985* 85.711*
ᵡ2/df 4.456 4.106 3.606
NFI .905 .914 .925
CFI .925 .933 .944
TLI .914 .923 .935
RMSEA .062 .059 .054
Based on Table 1, it was found that the 3-Factor Model which is the full measurement model under study was the best model that mostly fitted the samples data very well. This model was having the smallest χ², CMIN/DF and RMSEA. In addition, the hypothesized measurement model was the best fit since all of its goodness-of-fit were more than the required value and the highest among other model for fine-fitting model (i.e. NFI, CFI and TLI of > .90). Hence, this model was more superior as compared to other models. Furthermore, in assessing the extent to which a model exhibited the best fit, a univariate approach was used to determine if the difference in fit between the three models were statistically significant. As such, the differences in χ² (Δχ²) values between the three models was examined with the presumption that all models were nested and a significant Δχ² indicating substantial improvement in model fit. Comparison of 3-Factor Model with other models have yielded a differences in χ² of 151.985 if to compare with 1-Factor Model and differences of 85.711 if to compare with 2-Factor Model; while all the differences in χ² (Δχ²) were significant since p-value was < .000. Hence it was certain that all of these constructs in the full measurement model under study were having discriminant validity, thus it was confirmed that the scales that were developed to measure different constructs, were indeed measuring different constructs. 3.3 Full structural model Since the full measurement model was successfully conducted for each of constructs under this study, and it was satisfied that the measurement model was valid and well fitted to the data, thus it was best to proceed to full structural model to test the hypothesized research model between latent variables with full SEM. By executing the full structural model, then the hypothesised relationship between latent variables can be tested hence ascertaining the overall fit of the proposed model towards the samples data. In addition, once the full structural model was satisfied, the mediation effect shall be further testified against the alternatives. Based on the results of full measurement model and the proposed research framework as explained in the Methodology section, the full structural model was developed as per Figure 1. Based on Figure 1, certain constructs were modified during CFA and measurement model, and the relationship of each of the variables and its indicators were examined to test the hypotheses as well as to identify whether the proposed model fit the data well. As shown in Figure 1, satisfaction with work-life balance was hypothesised as partial mediating variables between work status congruence and Affective Commitment. Hence, this proposed model need to be verified against other alternative models i.e. Full Mediation Model (no direct relationship between work status congruence and Affective Commitment) and Non-Mediation Model (no direct relationship between affective commitment and Job Satisfaction). By accomplishing this model comparison verification, the best model shall be ascertained hence testifying the proposed partial mediation model and at the same time proving the whole model fit towards data sample. Based on results in estimates, it was found that all variables and their indicators were significantly related in a positive way. Hence, it is confirmed that there is a relationship between work status congruence and satisfaction with work-life balance hence supporting Hypotheses 1 with a significant standardised regression weight of .509. Similarly for Hypotheses 2, it is verified that there is a relationship between work status congruence and affective commitment with a significant standardised
10
Muhamad Khalil Omar / Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
regression weight of .253. As for Hypotheses 3, it was testified that there is a relationship between satisfaction with work-life balance and affective commitment with a significant standardised regression weight of .615.
e21 WLB7
1 1
Satisfaction with Work-Life Balance e1
e2 e3 e4 e5 e6
e7
1
1 1
1 1 1
1
WSC1
WSC4
1
WLB5
1
WLB4
1
WLB2
WSC3 Work Status Congruence
e14
WLB6
WLB3
WSC2
1
WLB1
e13 e12
e11
1
e10
1
e9
1
e8
WSC5 WSC6 WSC7
1
AC6 AC5
Affective Commitment
AC4 AC3
e22
AC2 AC1
1
e20
1 e19 1
e18
1 e17 1
1
e16 e15
Figure 1: Full structural model
Based on Table 2, it was found that the proposed Partial Mediation Model which is the full structural model under study was the best model that mostly fitted the samples data very well. This model was having the smallest χ², CMIN/DF and RMSEA. In addition, the hypothesised partial mediation model was the best fit since all of its goodness-of-fit were more than the required value and the highest among other model for fine-fitting model (i.e. NFI, CFI and TLI of > .90). Hence this model was more superior as compared to other models thus supporting Hypotheses 4 that the relationship between work status congruence and affective commitment is mediated by satisfaction with work-life balance. Furthermore, in assessing the extent to which a model exhibited the best fit, a univariate approach was used to determine if the difference in fit between the three models were statistically significant. As such, the differences in χ² (Δχ²) values between the three models was examined and in comparison with 2 other models have yielded a differences in χ² of 80.915 if to compare with Full Mediation Model and differences of 11.448 if
11
Muhamad Khalil Omar / Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
to compare with Non Mediation Model; while all the differences in χ² (Δχ²) were significant since p-value was .001. Table 2. Model fit summary (model comparison for mediation) Model Model 1 (Partial Mediation) Model 2 (Full Mediation) Model 3 (Non mediation) Differences (Model 3-1) Differences (Model 3-2) Note: * p≤0.001
ᵡ2 587.713 668.628 599.161
Df 163 164 164 1 1
Δᵡ2
80.915* 11.448*
ᵡ2/df 3.606 4.077 3.653
NFI .925 .914 .923
CFI .944 .934 .943
TLI .935 .934 .943
RMSEA .054 .059 .055
4. Conclusions This research has made its contributions with regard to the body of knowledge especially in extending the concept of work status congruence in work-life balance studies. Moreover, operationalisations of work status congruence and satisfaction with work-life balance were enhanced by inclusion of other nonwork activities and expanded on the subject of non-standard and temporary workers. Hence, this research has complemented prior studies of non-standard employment and work-life balance especially in substantiating positive impact of both domains towards affective commitment. Furthermore, effect of satisfaction with work-life balance as a mediator in work status congruence and affective commitment relationship has been established thus supporting previous scholars’ call for such confirmation. References Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: A review and recommend two-step approach. Psychological Bulletin, 103(3), 411-423. Aryee, S., Srinivas, E. S., & Tan, H. H. (2005). Rhythms of life: Antecedents and outcomes of work– family balance in employed parents. Journal of Applied Psychology, 90, 132–146. Baral, R., & Bhargava, S. (2010). Work-family enrichment as a mediator between organizational interventions for work-life balance and job outcomes. Journal of Managerial Psychology, 25(3), 274300. Byrne, B. M. (2010). Structural equation modeling with AMOS: Basic concepts, applications, and programming (2nd ed.). NY: Routledge Academic. Carr, J. C., Gregory, B. T., & Harris, S. G. (2010). Work status congruence’s relation to employee attitudes and behaviors: The moderating role of procedural justice. Journal of Business Psychology, 25, 583–592. DOI 10.1007/s10869-009-9151-z Connelly, C. E., Gallagher, D. G., & Gilley, K. M. (2007). Organizational and client commitment among contracted employees: A replication and extension with temporary workers. Journal of Vocational Behavior, 69, 326–335. Coyle-Shapiro, J. A-M., & Morrow, P. C. (2006). Organisational and client commitment among contracted employees. Journal of Vocational Behaviour, 68, 416-431. Fleetwood, S. (2007). Why work–life balance now? International Journal of Human Resource Management, 18(3), 387–400. Han, S.-S., Moon, S. J., & Yun, E. K. (2009). Empowerment, job satisfaction, and organizational commitment: Comparison of permanent and temporary nurses in Korea. Applied Nursing Research, 22, 15–20. Higgins, C., Duxbury, L., & Johnson, K. L. (2000). Part-time work for women: Does it really help balance work and family? Human Resource Management, 39(1), 17–32.
12
Muhamad Khalil Omar / Procedia - Social and Behavioral Sciences 107 (2013) 4 – 12
Hill, E. J., Ferris, M., & Martinson, V. (2003). Does it matter where you work? a comparison of how three work venues (traditional office, virtual office, and home office) influence aspects of work and personal/family life. Journal of Vocational Behavior, 63(2), 220–241. Holtom, B. C., Lee, T. W., & Tidd, S. T. (2002). The relationship between work status congruence and work-related attitudes and behaviors. Journal of Applied Psychology, 87(5), 903–915. Hyman, J., & Summers, J. (2004). Lacking balance? Work-life employment practices in the modern economy. Personnel Review, 33(4), 418-429. Jang, S. J., Park, R., & Zippay, A. (2011). The interaction effects of scheduling control and work–life balance programs on job satisfaction and mental health. International Journal Social Welfare, 20, 135–143. DOI: 10.1111/j.1468-2397.2010.00739.x Martin, J. E., & Sinclair, R. R. (2007). A typology of the part-time workforce: Differences on job attitudes and turnover. Journal of Occupational and Organizational Psychology, 80(2), 301–319. McNall, L. A., Masuda, A. D., & Nicklin, J. M. (2010). Flexible work arrangements, job satisfaction, and turnover intentions: The mediating role of work-to-family enrichment. The Journal of Psychology, 144(1), 61–81. Meyer, J. P., Allen, N. J., & Smith, C. A. (1993). Commitment to organizations and occupations: extension and test of a three-component conceptualization. Journal of Applied Psychology, 78, 538– 551. Moore, F. (2007). Work-life balance: contrasting managers and workers in an MNC. Employee Relations, 29(4), 385-399. Pearce, J. L. (1993). Toward an organizational behavior of contract laborers: Their psychological involvement and effects on employee coworkers. Academy of Management Journal, 36, 1082–1096. Polivka, A. E., & Nardone, T. (1989). The definition of contingent work. Monthly Labor Review, 112, 916. Sturges, J., & Guest, D. (2004). Working to live or living to work? Work/life balance early in the career. Human Resource Management Journal, 14(4), 5-20. Thorsteinson, T. J. (2003). Job attitudes of part-time versus full-time workers: a meta analytic review. Journal of Occupational and Organizational Psychology, 76, 151-77. Valcour, M. (2007). Work-based resources as moderators of the relationship between work hours and satisfaction with work–family balance. Journal of Applied Psychology, 92(6), 1512–1523. Van Dyne, L., & Ang, S. (1998). Organizational citizenship behavior of contingent workers in Singapore. Academy of Management Journal, 41, 692–703. Walker, B. (2011). How does non-standard employment affect workers? A consideration of the evidence. New Zealand Journal of Employment Relations, 36(3), 14-29. Wayne, J. H., Randel, A. E., & Stevens, J. (2006). The role of identity and work–family support in work– family enrichment and its work-related consequences. Journal of Vocational Behaviour, 69(3), 445461.