The predictors and performance-related outcomes of bi-directional work-family conflict : An empirical study

The purpose of this paper is to investigate the relationship of the conflicting domains of work and life with employees' self-evaluated performance. Absenteeism has also been included in the study as a variable that has direct effects on both family-work conflict (FWC) and performance. Demographic variables have been taken in as background predictors of the independent variables of the basic model. The study was conducted on employees of local banks in Rawalpindi/Islamabad. Findings revealed that work-family conflict (WFC) has no significant effect on performance whereas FWC and absenteeism were found to be significant and negative predictors of performance. Absenteeism was also found to have a positive association with FWC. A cultural analysis was conducted to understand the effects of the demographic variables.


INTRODUCTION
Over the past few decades, with the changing life patterns globally, and due to the constant shift towards a service economy (Grzywacz et al., 2007), a new domain of balancing professional and personal lives has arisen.Work-family balance is defined as the integration of work demands with family roles (Gropel and Kuhl, 2009).The conflictive association of work and family is a potential cause of stress and deviant work attitudes (Schieman et al., 2003).
With the evolution of bi-directional work-family balance issues like work-family conflict (WFC) and family workconflict (FWC), the managers and decision-makers are concerned to know if this concept actually affects the quality of performance of an employee.Moreover, other psychological and job-related variables also play their part in determining the strength of this causality.It is therefore a challenge for researchers to identify the antecedents, moderators and results of this bi-directional causality (Rotondo and Kinsaid, 2008).Time is one important factor that determines the level of conflict in an individual's schedules.The way an employee manages time is an important determinant of his/her level of work-life balance.Had it been reserved to the level of flexibility on the employer's behalf, the issues of work-life imbalance would have been long forgotten due to the interventions that organizations attempt in form of family-friendly workschedules.
Research identifies that the individual plays an imperative part in determining his/her own level of workfamily balance (Rotondo and Kinsaid, 2008).However, there are other external forces that affect the individual in doing so.Keeping those variables out of the scope of this research, the basic focus is laid upon individualistic jobrelated deviant behaviors -particularly absenteeism.
These stressors, however, do not remain similar through all phases and stages and conditions of human life.Demographic variables like age, gender, marital status, number of children and education do affect these performance predictors.This study focuses on the root causes of these stressors and their consequences on performance.

Rationale
The aim of this research is to study the individual aspect of work-life conflict, the effects on self-evaluated performance and productivity as well as the role of absenteeism and family-work conflict in this process.The domain of work-life balance has been under study for the last 4 to 5 decades; however, no progress has been done on a remedial basis.The causal relationship is still under study and no worthwhile pattern has yet been identified.Many of the efforts on organizational basis have failed to solve this issue.Most of the research is carried out on the consequences of work-family conflict instead of its causes or even its very definition (Greenhaus et al., 2002).This study aims at providing a different insight to the literature from the individual's perspective so that remedial measures are also taken in this direction.This study is an initial effort of studying work-life balance as a personal phenomenon; particularly in this part of the world.Moreover, the study aims at testing the Western concepts of work-family and family-work conflict in a society with a different concept of both family and work.

LITERATURE REVIEW
Work-family balance is a subjective concept based upon perception of availability of time for work and family domains (Gropel and Kuhl, 2009).It is a perceived state that an individual believes he or she is in which affects and individual's work and life attitudes, subjective wellbeing as well as the level of stress.As this research studies the bi-directional variable from both dimensions, the first consideration is work-family conflict.It is identified, by the literature, as a poignant stressor that may result into several health-related outcomes, thus reducing the level of performance (Grzywacz et al., 2007).Work-family conflict has been identified as a negative force that has negative job-related outcomes like reduced performance and productivity levels.
More precisely, it has been defined as a clash between roles, each with certain "situational urgencies", hindering performance in each other (Greenhaus et al., 2003).It causes a loss of productivity (Schieman et al., 2003).Based on the support from previous studies, this research hypothesizes that: H 1 : Work-family conflict has a negative impact on employee's self-evaluated performance.
Absenteeism has been included in the study, as to get a clear picture of where it stands in the WFC, FWCperformance relationship.In some studies, it has been referred to as a consequence of family-work conflict (Anderson et al., 2002;Schieman et al., 2003).Absenteeism has been studied in stress-related researches as a consequence of job-dissatisfaction, Mohsin and Zahid 11505 unfairness etc. which are performance-reducing variables.Absenteeism itself has also been identified by the literature as a source of reduction in performance (Marburger, 2001).The study has therefore, hypothesized that: H 2 : Absenteeism has a negative impact on employee's self-evaluated performance.
Family-work conflict is the interference of a worker's role as a family-member with his/her role at job.This is a potential stressor (Anderson et al., 2002), which also affects performance adversely (Eby et al., 2005).In a meta-analysis, family-work conflict was found out to be more intensely a cause of reduction in employee performance than work-family conflict (Kossek and Ozeki, 1999).It has been hypothesized that: H 3 : Family-work conflict has a negative impact on employee's self-evaluated performance.H 4 : Absenteeism has a positive impact on FWC.

Background variables
The research does not only focus on the performancerelated outcomes of the bi-direction work-family conflict; it also focuses on the causes that result into these situations.Therefore, other than the main research variables, background variables were also included in the study -age, gender, marital status, number of children, level of education and tenure in organization.Certain hypotheses were drawn regarding the effects of these variables on the basic hypothetical model.Positive effects of age, education and tenure are expected on WFC.All these three assumptions are supported by literature (Cohen and Liani, 2008;Gordon et al., 2007): H 5 : WFC is positively affected by age H 6 : WFC is positively affected by education H 7 : WFC is positively affected by tenure Moreover, positive effects of marital status and number of children are expected on FWC (Cohen and Liani, 2008;Gordon et al., 2007).However, these are variables have different effects across different cultures.Most of the literature that this research has borrowed from has been experimented in countries other than that under study with different cultures and expected responsibilities: H 8 : FWC is positively affected by marital status H 9 : FWC is positively affected by number of children H 10 : FWC is positively affected by gender (females have a greater level of FWC as compared to males) Tenure in organization is also expected to be a determinant of absenteeism due to security of job and adjustment with peers (Cohen and Liani, 2008).Similarly, studies have also shown that age is a positive determinant of absenteeism due to health issues: H 11 : Tenure has a positive impact on absenteeism H 12 : Age has a positive impact on absenteeism

Model
Figure 1 shows the conceptual model that has been established on the basis of the literature review.

METHODOLOGY Population
The population for the study included the employees of the banking sector in Rawalpindi and Islamabad.This population was chosen because of the long working hours and tough schedules causing a shortage of time available for family.This indicated, to the researchers, an existence of work-family conflict issues.The estimated size of the population is below 10,000.

Sampling
The stratified random sampling method was used.Five local banks were selected and a total of 30 employees from each bank were included in the study.Only the local banks were included in the study to avoid distortions in study due to cultural variances.The sample size with a 95% confidence level was calculated to be 350.

Instruments
Work-family conflict and family-work conflict were measured using a 6-item questionnaire developed by Russell et al. (2010).The questionnaire consisted of 3-item for each of the variables with a 5point Likert scale ranging from strongly agree to strongly disagree.Both variables were measured upon the same components: strain, behavior and time.
Employee performance was measured using a self-appraisal questionnaire by Yousef (2000).The 4-item questionnaire was based on a 7-point Likert scale ranging from very low to very high.The facets that measured performance were productivity and performance.
Absenteeism was measured using a 3-item questionnaire with the following components: frequency, preference and ease.The questionnaire was based on a 5-point Likert scale.

Tools
The results were analyzed using correlation and hierarchical regression using appropriate statistical package.

RESULTS
Table 1 shows the results obtained after the analysis of the data.The relevant significant correlations have been highlighted.Discussing the basic four hypotheses, the correlation between WFC and performance does not stand very significant.However, corresponding to the other two hypotheses, there is a significant negative correlation of -0.23 between FWC and performance and a strong and significant negative correlation of -0.41 between absenteeism and performance.Similarly, the relation between FWC and absenteeism is also significant with a value of 0.32.Among the background variables, the relationships of WFC with age, education and tenure have all not been very significant in these results.The determinants of FWC to some extents, have met the expectations of the research.Though marital status and gender do not yield significant results; the relation between FWC and number of children has a strong and positive relation of 0.47 with FWC.Un-hypothesized though, there can be seen a significant positive relationship between gender and absenteeism.Among the hypothesized relations, only the effect of tenure in organization shows a significant positive relation of 0.17 with absenteeism.The effect of age is not significant.

Regression
The first regression model depicts the first 3 hypotheses that correspond to the effects of WFC, FWC and absenteeism upon performance (Table 2).
The Beta values of FWC and absenteeism show negative slopes in relation to performance -FWC has a steeper slope of -0.112, whereas absenteeism has a flatter slope of -0.373 which shows a higher rate of responsiveness of performance.No pattern of causality has been identified between WFC and performance.
The second regression model explains the causalities between WFC and age, education and tenure.The model explains 1% change in WFC due to the 3 predictor  variables, which is not a very significant causality.The overall strength of the model is also not supported by the F-statistic, which is insignificant upon a 95% confidence level.
The third regression model illustrates the effects of marital status, number of children and gender on FWC.The significant R 2 of 23.2% depicts that this percentage of change in FWC is explained by the proposed model.However, we separately study the effect of each variable on FWC.
Tables 3 and 4 shows a strong model with a significant F value.However, the beta values figure out that it is due to the strong positive causality between number of children and FWC that explains the high level of Rsquare and the other two predictor variables do not have very significant effects.
The fourth regression model corresponds to the effects of tenure and age of absenteeism (Table 5).With 3% explained variation, though this is not a very strong model; the positive effect of tenure on absenteeism can be identified.The effect of age, however, is not explained by this model.

DISCUSSION
It has been identified that no hypotheses corresponding to work-family conflict, has been accepted; whereas, literature supports all these hypotheses, including its effect on performance, and the effects of age, education and tenure on itself.It has been observed in the data that WFC, other than being a form of inter-role conflict, has adopted the face of a lamenting behavior that all employees carry.Every employee, regardless of age, gender, education, tenure and performance level, believes that he/she has to work more than he/she actually should and therefore the data is highly skewed towards the right.And both positive and negative hypotheses stand meaningless.Family-work conflict is yet to be realized as a form of conflict in the society under study where family is a strong institution and any responsibilities corresponding to it are regarded as essential, and any amount of time or form of behavior required in family is referred to as the individual's dominant behaviors.Work is yet to be identified as an equal if not higher priority to family; therefore, this form of conflict is not very well comprehended.Moreover, the strong institution is also supportive in nature and shortcomings in responsibilities towards it are adjusted and the work requirements of the individual are also understood by family.Therefore, the individual does not realize, or is not made to realize this form of inter-role conflict.Nonetheless, among those who realize this conflict are working parents whose family time coincides with work time.That is why the hypotheses referring to the effect of number of children on FWC is highly significant.
The effects of age on absenteeism have not been identified for the reason that the basic assumption that this hypothesis rested upon was the health issues relating to old age; conversely, none or very few of the employees interviewed were of a very high age as to have serious health concerns.The hypothesis related to tenure can be explained by the adjusting behavior of the society under study.It is common culture that people are socially flexible and adjust each other's inadequacies especially when the social bond matures with time and age.People who work in organizations for long, in this society, believe they are understood and can take advantage of it.Therefore, the effect of tenure on absenteeism is positive.

Conclusion
Based upon these findings, no effect of WFC has been identified on performance, therefore H 1 is rejected.However, the effects of FWC and absenteeism are significant and H 2 and H 3 are accepted.FWC is also predicted significantly by absenteeism, therefore H 4 is also accepted.
No effects of age, education and tenure have been identified on WFC; thus H 5 , H 6 and H 7 are rejected.No particular trend has been identified in the relationship of FWC with marital status and gender.Thus, H 8 and H 10 are also rejected and it is concluded that FWC is not affected by marital status and gender.However, there is a very strong effect of number of children on FWC, so H 9 is accepted and it is concluded that employees with a larger number of children have a higher degree of FWC.
The effect of tenure on absenteeism is also significant; thus H 11 is accepted.The effect of age on absenteeism is though not recognized and H 12 is rejected.

RECOMMENDATIONS
For future research, this study paves way for the future studies to investigate this domain.It is recommended that coming studies should include other deviant work behaviors as mediator variables between the forms of conflict, which this study could not pursue due to temporal delimitations.These variables may include turnover intentions, dishonest practices, workplace violence, etc.
Moreover, it is recommended that these psychological variables should also be studied through ethnographic tools as to get a clearer picture of the prevailing cultural predictors and historic trends.Anthropological methods of researching the work practices can bring up better and deeper insights into to the cause and effect relations of these variables, which will results into more profound understanding of the subject.

Table 1 .
Results of data analysis.

Table 2 .
Analysis of the first regression model.

Table 3 .
Analysis of the second regression model.

Table 4 .
of the third regression model.

Table 5 .
Analysis of the fourth regression model.
b Dependent variable: Ab.