Women, work & well-being: the influence of work-family and family-work conflict.
Following international trends, the structure of New Zealands traditional households is experiencing some significant changes. For example, in the ten years between 1986-1996 a 2% increase in the number of single person dwellings in New Zealand was observed (Statistics New Zealand, 2001). There was also an increased number of single parent households and this trend is expected to continue (Amato, 2000). Projected figures suggest that by the year 2021 there will be significantly fewer children under the age of 14 years living in New Zealand, as compared to current figures (Statistics New Zealand, 2001). These demographic trends appear to suggest that both men and women are now choosing to remain single for a longer period of time, and the decision to have children is being increasingly deferred or put-off completely (Amato, 2000).
Changes in household structure are occurring partly as a result of recent employment developments, most noticeably the emergence of dual-career partnerships and increased opportunities for female career participation within the workplace (Rudman, 1999; Statistics New Zealand, 2001). These developments are expected to continue as more women attain tertiary qualifications (Eagle, Miles & Icenogle, 1997; O'Driscoll, 1996). For most women however, their participation in paid employment has not meant the abandonment of their traditional familial role as the primary caregiver to children and other dependents (Hammer, Allen & Grigsby, 1997; Kline & Cowan, 1988). Indeed, it has been suggested on numerous occasions that women generally continue to spend more time working within the household as compared to their male counterparts (e.g. Burke & McKeen, 1992; Parasuraman, Greenhaus & Granrose, 1992; Skues, & Kirkby, 1995). For example, Hochschild (1989) found that female advancement in terms of equal opportunities within paid employment was not matched by similar advances of equality within the home life. Hochschild estimated that the additional demands experienced by women in dual-career families amounted to a `second shift' of work or an additional working month of 24 hour days per annum, as compared to men within a similar situation. Gjerdingen, McGovern, Bekker, Lundberg, & Willemsen (2000) also recorded similar findings within an international study, comparing working loads across Sweden, the Netherlands and the U.S.
The effects of both work and non-work (family) roles upon individual psychological well-being have recently been examined within New Zealand and other countries (O'Driscoll, 1996; O'Driscoll & Humphries, 1994). It is apparent that the isolation of an individual's work and non-work roles can no longer be supported in psychological well-being measurement, instead attention has turned to the various interactional pathways between these two domains (e.g. Greenhaus & Beutell, 1985). The interdependence of an individual's work and non-work roles and especially the conflicting demands between these roles, has become known as inter-domain conflict: "... in which the role pressures from the work and family domains are mutually incompatible in some respect" (Kahn, Wolfe, Quinn, Snoek & Rosenthal, 1964, cited in Greenhaus & Beutell, 1985, p.77). Note that the use of the term inter-domain conflict within this paper follows this generic association and thus incorporates conflict in two directions: both work[right arrow] family and family[right arrow]work conflict. Only where necessary is the specific direction of the conflict explicitly stated.
It has been suggested that there are three major forms of inter-domain conflict; time-based conflict, strain-based conflict, and behaviour-based conflict (e.g. Carlson, Kacmar & Williams, 2000; Pleck, Staines & Lang, 1980). Time-based conflict refers to the individual's inability to perform a task in one role, either physically or due to a cognitive preoccupation, as a result of time demands in another role (Greenhaus & Beutell, 1985). For example, an approaching work deadline may result in more time spent at work, and thus less time with a partner/family. Similarly, high levels of personal/family demands may also contribute to absenteeism from the workplace. Strain-based conflict is experienced when a pre-occupation with the demands from one role, interferes with the ability to perform adequately within a second role. Thus for example, the anxiety caused by an approaching work deadline may result in the `carry-over' of stress reactions to the home domain, adversely influencing performance within this second domain. Finally, behaviour-based conflict refers to the display of specific behaviours in one domain which are incongruous with the desired attributes of the second domain. Thus for example, a supportive and loving parent may exhibit contrasting behaviour when involved with a hostile work situation. The interested reader is directed to Greenhaus & Beutell, (1985) for a more detailed explanations of these constructs.
The basic models of inter-domain conflict argue that stressors from one role (e.g. work) have adverse `strain' effects on the individual in their other role (e.g. family/nonwork) based on any one, or a combination of these three forms of inter-domain conflict. Furthermore, most research in this area has suggested that women generally report greater inter-domain conflict than their male counterparts, primarily due to high levels of family demands (Aryee & Luk, 1996; Aryee, Luk, Leung & Lo, 1999; Hammer et al. 1997).
The nature of this inter-domain conflict and the factors which influence its occurrence and intensity have recently been explored. For example, it has been found that working women with young children suffer greater inter-domain conflict and consequently more adverse outcomes than working women with older children (Rosenbaum & Cohen, 1999; Thompson, Beauvais & Lyness, 1999). Eagle et al. (1997) suggest that the more children there are in the family per se, the greater the inter-domain conflict of the parents. In addition, the disadvantages of being a lone parent in relation to psychological well-being have also been documented (e.g. Baker & North, 1999). Increased bi-directional inter-domain conflict (i.e. both work [right arrow] family and family[right arrow]work conflict) has been found to be greater amongst women who have partners than among single women (Herman & Gyllstrom, 1977), although this finding contradicts the more widely acknowledged suggestion that marital status is considered to be a positive contributor to well-being, and especially so for men rather than women (Coombs, 1991). However, research on the positive influence of marital status has produced inconsistent findings and in some cases has not been fully replicated (see for example: Brown, 2000).
As well as demographic variables, a number of dispositional and situational factors which influence inter-domain conflict have also been explored. The influence of perceived control and perceived social support are of particular interest here. The experience of high levels of these two variables within the work domain has been identified as offering some protection from the negative effects of occupational stress (e.g. Lazarus & Folkman, 1984; Parasuraman, Purohit & Godshalk, 1996; Pearlin & Schooler, 1978). For example, in a review of the literature evaluating Karasek's (1979) Job Demand-Control Model, Van der Doer and Maes (1999) concluded the research indicates that both perceived job-related control and perceived job-related social support positively affect the nature of the stressor-strain relationship within a work context (i.e. through a buffering or moderation role), such that the adverse consequences of increased work demands can be reduced (moderated) through the perception of high levels of job control and/or high levels of perceived work-based social support.
When investigating psychological well-being, the influence of these third variables (such as perceived control and social support) should routinely be taken into account. The presence of such third variables has been found to significantly influence well-being outcomes, and the need to measure and control for this influence in order to accurately predict outcomes has been acknowledged (e.g. Baron & Kenny, 1986; Oliver & Brough, in press; Zapf, Dormann, & Frese, 1996).
Inter-domain conflict has also been found to influence psychological well-being and has a direct influence on the relationship between the presence of a stressor and the experience of strain (Frone, Russell & Cooper, 1992). However, the presence of additional third variables within this relationship, such as various individual demographic circumstances, has also been recently acknowledged (Burke & Greenglass, 2001; Voydanoff, 2002). For example, when measuring inter-domain conflict, the presence/absence of a partner and/or dependents has been found to explain variations within such constructs as individual perceptions of control (Rosenbaum & Cohen, 1999) and social support (Nielson, Carlson, & Lankau, 2001), as well as having direct effects upon psychological well-being (Richardsen, Burke & Mikkelsen, 1999).
This paper intends to examine these relationships in further detail. Demographic characteristics and measures of both perceived control and social support are evaluated in terms of experienced inter-domain conflict. The influence of perceived control and social support from each domain (job and family) will be measured. The ability of each of these variables to predict psychological well-being is estimated. A sample of New Zealand women with various demographic characteristics were invited to participate in this research. Specifically, women fitting the demographic criteria were recruited: those with and without partners and those with and without dependents, thus producing four distinct sample groups. Comparisons between these four demographic groups were conducted, to test the following hypotheses:
Hypothesis 1: The influence of dependents
It is hypothesised that the respondents who have dependents (regardless of having a partner or not) will report greater levels of both work [right arrow] family (H1a) and family [right arrow] work (H1b) conflict than respondents with no dependents. Thus the presence of dependents will produce higher levels of inter-domain conflict. It is suggested that high levels of inter-domain conflict will in turn, be associated with adverse psychological well-being outcomes (H1c).
Hypothesis 2: The influence of a partner: control
It is also hypothesised that respondents without a partner (regardless of the presence of dependents)will have higher perceived levels of control within their non-job environments (H2a) than respondents who do have a partner. Thus, being single will provide the opportunity for higher levels of perceived (family) control (i.e. self dependency rather than dissipated responsibility, will result in higher levels of perceived individual control at home). High levels of control will in turn, be associated with positive psychological well-being outcomes (H2b).
Hypothesis 3: The influence of a partner: social support
Finally, it is suggested that the respondents who do have a partner (regardless of the presence of dependents) will report more family social support than respondents without a partner (H3a). Thus the presence of a partner will contribute to a greater perception of increased family social support. High levels of social support will in turn, be associated with positive psychological well-being (H3b).
Participants and Procedure
A total of approximately 75 large organizations in New Zealand were invited to participate in this research. Of this number, 23 organisations actually participated, producing a total participant pool of 1495 employees. In addition, 300 members of the New Zealand Institute of Management (NZIM) were also invited to participate. The participating organisations represented a wide diversity of industries, including: financial, retail, manufacturing, tourism and service industries. The Human Resources officer (or equivalent) in each organisation provided contact details of a sample of approximately 50-100 employees for research participation. As far as was possible, each sample consisted of an approximate equal mix of genders, ethnicities, organisational levels and geographic locations according to the characteristics of each specific organisation.
The 1795 participants were required to complete two confidential self-completion questionnaires, administered at a three-month interval. The responses were returned directly to the researchers in a reply-paid envelope. A total of 691 employees responded to the first questionnaire, constituting a response rate of 38.5%. Of this number, 371 respondents were female (53.6%). This paper reports the results for the female respondents to the Time 1 questionnaire only.
The majority of the respondents (86%) were of Pakeha/New Zealand European descent. Maori constituted 7% of the sample. The respondents ranged in age from 21-67 years, with an average of 37 years. More than three quarters of the respondents had a partner (78%), and the majority of these respondents lived with their partner all of the time (74%). Just under half of the respondents had a/some dependent(s) (49%), the vast majority of whom were children (92.5%). Overall, 41% of the respondents had both a partner and dependent/s (family group), 38% had a partner only (partner group), 13% had neither partner and dependent/s (single group) and 9% had dependent/s only (solo carers group). The respondents with dependent children signified that they were responsible for between 1-9 children, with an average of 2. The age of these children ranged between 4 months to 27 years. Of the 28 dependents who were not children, these were identified as consisting of: brothers, sisters, aunts, uncles, parents, grandchildren, cousins and various in-laws. Bachelors and higher degrees were held by 28% of the respondents.
The majority of the respondents were employed in the professional and business sector, in particular the banking, legal, science and technology industries, as well as the NZIM (73%). The remaining respondents were employees of various service and manufacturing firms. The average length of current job tenure was 4 years. The respondents worked an average of 41 hours per week. Of the 268 respondents who indicated that they had a partner, 91% of these respondents indicated that their partner was employed and worked an average of 44 hours per week.
Comparisons between these demographic variables were conducted and the results were found to be within the generally expected directions. Thus for example, the single respondents tended to be the youngest respondents. The solo carer respondents were employed for significantly fewer hours per week (mean of 36 hours) as compared to all the other respondents (mean of 43 hours). Additionally, the respondents who had both a partner and dependents were employed for significantly fewer hours per week (mean of 40 hours), as compared to those respondents who had a partner and no dependent(s) (mean of 43 hours). Thus, the respondents with dependents were found to work fewer hours in paid employment each week, as compared to the respondents without any dependents.
Inter-domain conflict: Work[right arrow]family conflict (WFC) and family[right arrow]work conflict (FWC) were measured by two five item measures of inter-domain conflict (Netemeyer, Boles and McMurrian, 1996). The WFC measure includes such items as: `The demands of my work interfere with my home and family life' and `My job produces strain that makes it difficult to fulfil family duties'. Acceptable internal reliability coefficient (Cronbach's alpha) have been reported for this measure: .88 (Netemeyer et al. 1996). The FWC measure includes such items as: `I have to put off doing things at work because of demands on my time at home' and `Family-related strain interferes with my ability to perform job-related duties'. Acceptable internal reliability coefficient (Cronbach's alpha) have also been reported for this measure: .89 (Netemeyer et al., 1996). For both measures, the respondents indicated their agreement with each item on a seven-point Likert scale (1 = `strongly disagree' to 7 = `strongly agree'). High scores thus indicate high levels of conflict.
Perceived control: Perceived individual control was measured by an adaptation of Pearlin and Schooler's (1978) seven item Mastery scale. This measure was included both as a family related measure and as a job related measure. The respondents were asked to complete both measures. An item from the family-related control measure reads: `I often feel helpless in dealing with the problems of family life'. The original (domain-free) measure has acceptable internal reliability coefficient (Cronbach's alpha) of .80 (Thoits, 1987). The use of this perceived control measure in a specific family related direction has not previously been reported. An item from the job-related control measure reads: `I have little control over the things that happen to me in my job'. A moderately acceptable internal reliability coefficient (Cronbach's alpha) of .60 has been reported for this measure (Brough, 1997). Both measures were answered according to agreement with each item on a seven-point Liken scale (1 = `strongly disagree' to 7 = `strongly agree'). High scores thus indicate high levels of perceived control.
Social support: A measure of perceived social support developed in an earlier version of this research was included here (O'Driscoll, 2000). The measure requires the respondents to identify the frequency of receiving different kinds of support from both work colleagues (four items) and from family members (four items). The four types of social support measured were `helpful advice or information', `sympathetic understanding and concern', `clear and helpful feedback' and `practical assistance'. The respondents scored each item according to a six-point Likert scale (1 = `never' to 6 = `all of the time').
Psychological strain: Context-free psychological well-being was measured with the 12 item General Health Questionnaire (GHQ-12; Goldberg, 1972). The GHQ was designed to identify mental health symptoms and is widely used in both clinical and occupational settings (for example: Banks, Clegg, Jackson, Kemp, Stafford & Wall, 1980; Parker, Wall & Myers, 1995). The respondents were asked to evaluate their psychological well-being in the context of the previous three months, according to such items as: `Lost much sleep over worry' and `Felt constantly under strain'. Pursuant to O'Driscoll (2000), a six-point Likert scale was employed, with responses ranging from 0 = `never' to 5 = `all the time'. High scores indicate high levels of psychological strain. O'Driscoll (2000) reports an acceptable internal reliability coefficient (Cronbach's alpha) of .88 for this measure.
Demographic variables: The respondents were asked to provide information regarding their gender, ethnicity, age, marital status, living arrangements, number of children, the ages of these children and the number and relationship to any other dependents who lived with them. The respondents also answered questions concerning the highest level of education they had completed, their current job title and tenure, and the number of hours they and their partner (where applicable) worked each week. Note the gender of the partner was not requested in this research.
Prior to the statistical analyses, the data were examined for extreme (outlining) cases and these were excluded from all subsequent analyses (N=344). The data were also verified against the normal statistical assumptions. The unequal sample sizes in the four respondent groups were noted. It is assumed here that these differences largely reflect the demographic characteristics of the general population, i.e within the general population more individuals reside in family groups as compared to single households (Statistics New Zealand, 2001; Tabachnick & Fidell, 1996). The distribution of the research variables was assessed within each of the four respondents groups. Although producing unequal numbers, the number of cases to variables ratio for each cell did not violate the statistical assumptions. Furthermore, although a correlation was observed between the two variables of partner and dependents, upon which the respondents groups are formed, this correlation was small (r= .13, p<.01) and therefore was considered to be acceptable for this categorisation (Tabachnick & Fidell, 1996). The respondents were therefore divided into four categories for the purposes of the multivariate analyses: single (no partner and no dependents), solo (dependents only), partner (partner only) and family (both partner and dependents).
Descriptive statistics and correlations
The scale descriptives for the research variables are illustrated on Table 1. The means, standard deviations and alpha coefficients for the scales are acceptable and are generally comparable with the statistics reported in the literature. One interesting finding concerns the difference in mean values between the two inter-domain conflict variables. With a maximum range of 35, the work [right arrow] family variable produced a considerably higher mean (of 20.44) as compared to the family [right arrow] work variable (12.43). This implies that the respondents generally perceived their work demands as adversely interfering with their home-life to a greater extent than family to work interference.
The respondents with a partner reported significantly higher levels of family control, although this association was small (r =. 18, p<.001). The presence of a partner was also negatively associated psychological strain (r= -.1 [right arrow] 12, p<.05). No association was identified between the presence of a partner and either type of inter-domain conflict. The presence of dependents however, was significantly associated with both types of inter-domain conflict: work [right arrow] family conflict (r = -.15, p<.01) and family [right arrow] work conflict (r = -.23, p<.001). No significant relationship was identified between dependents and psychological strain. Both types of inter-domain conflict were significantly associated with psychological strain, with work[right arrow]family conflict producing the stronger association (r = .27, p<.001). Finally, both domains of perceived control and family social support were all found to be significantly associated with decreased psychological strain. Perceived job-related control in particular, produced a strong relationship with this criterion (r =-.48, p<.001).
Differences in mean scores between the four respondents categories were tested with the General Linear Model (GLM) procedure, which is largely similar to the MANOVA procedure. The overall GLM F was significant on all levels (Pillai's trace, Wilks' lambda, Hotelling's trace and Roy's largest root, all p<.001) suggesting progression with the univariate analysis should proceed. The results of this univariate analysis are presented in Table 2. Descriptive statistics for the significant univariate associations generated by the GLM are provided in Table 3.
It can be observed that both types of inter-domain conflict differed significantly across the four respondents groups (Table 2). An examination of the descriptive statistics (Table 3) reveals that both the solo and family groups produced significantly higher mean scores for family [right arrow] work conflict and additionally, the solo group produced the highest mean score for work[right arrow]family conflict. These results imply that the presence of dependents is associated with increased conflict, and this is especially strong for family [right arrow] work conflict. These differences between the respondents groups are illustrated in Figures 1 and 2.
[FIGURES 1-2 OMITTED]
The respondents groups also differed on their levels of family control. Both the single and solo respondents groups produced significantly lower mean scores compared to their counterparts. These results imply that the presence of a partner is associated with increased perceived levels of family control. This result is illustrated on Figure 3. Finally, a difference in psychological strain was also identified between the respondents groups. The single respondents produced the highest mean scores, indicative of experiencing more psychological strain, as compared to their colleagues (see Figure 4).
[FIGURES 3-4 OMITTED]
To test the influence of the demographic and psychological variables upon psychological strain a hierarchical multiple regression was conducted and is displayed in Table 4. The demographic variables (partner and dependents) were entered on the first step and contributed minimally to the variance explained within the criterion ([R.sup.2] = .02, F(2, 316) = 3.51, p<.05). At step two, the predictors relating to the family situation were entered (work [right arrow] family conflict, family social support and family control), in order to determine the effects of the family domain upon the criterion. Both work [right arrow] family conflict and family control accounted for significant proportions of individual variance (25% and -35% respectively, both p<.001). In combination, the family domain predictors explained 21% of the final shared variance within the criterion (F(5, 313) = 19.08, p<.001).
The job related predictors (family [right arrow] work conflict, perceived job-related social support and control) were entered into the equation at the third and final step. The conflict and control predictors again accounted for significant proportions of unique variance within the criterion (11%, p<.05; -40%, p<.001, respectively). In combination, the three job related predictors contributed an additional 14% (p<.001) of explained variance. Job related perceived control contributed the largest proportion of individual variance within this equation. The regression equation explained 37% of the variance occurring within psychological strain (F(8, 310) = 22.94, p<.001).
The influence of dependents
The respondents with dependent(s) reported higher levels of inter-domain conflict, with a stronger association produced for family [right arrow] work conflict. Strong support for hypothesis (H1b) was therefore obtained: dependents increase perceived levels of family [right arrow] work conflict. This result reflects the findings of previous studies which also concluded that the demands produced by dependents do adversely affect work tasks and performance (e.g. Greenhaus & Beutell, 1985; Rosenbaum & Cohen, 1999; Thompson et al. 1999). Less support was obtained for Hypothesis (H1a) in the current research. Only one group of respondents with dependents (solo carers) reported significantly higher levels of work [right arrow] family conflict. This finding was not replicated with the other respondents who also had dependents (and a partner). Increased experiences of work [right arrow] family conflict are therefore, not only due to the presence of dependents.
Both types of inter-domain conflict were associated with, and predictive of, increased psychological strain, supporting Hypothesis (H1c). The conflict variables each individually predicted approximately 12% of variance within psychological strain. This result supports other findings of this nature. The adverse impact of inter-domain conflict upon psychological well-being appears to be firmly established within the literature (Baker & North, 1999; O'Driscoll, 1996). This paper has illustrated that both pathways of inter-domain conflict adversely affect psychological well-being to a similar extent.
Interestingly, this research revealed a positive relationship occurring between dependents and psychological strain. In the final step of the multiple regression equation dependents explained 10% (p<.05) of the unique variance within the criterion. Other researchers have also suggested that the presence of dependents (children) is associated with adverse psychological outcomes (e.g. Eagle et al. 1997). Although not explored in detail here, it is expected that the type, age and health of the dependent(s) are likely to influence this relationship. This association is becoming increasingly important to define accurately as the demographic forecasts suggest that ageing parents, rather than developing children, will account for increasing proportions of the dependent population (Statistics New Zealand, 2001).
No similar relationship was observed for the partner predictor variable with psychological strain, although the GLM analysis did reveal that the single respondents achieved the significantly highest mean score within the psychological strain variable. Similarly, the correlation analysis indicated a small but significant, negative association existed between these two variables, suggesting that the presence of partner was associated with increased well-being. Further analysis of the larger data set will indicated whether this finding can be replicated for the male respondents also. The relationship between marital status and psychological well-being has been recently discussed elsewhere (e.g. Brown, 2000). The mixed findings produced by these recent investigations largely reflect the findings of the current research. The presence of a partner does appear to be associated with improved well-being, however any such predictive influence is removed when other factors are simultaneously considered.
The influence of a partner: control
In a reversal of the hypothesised findings, those respondents with a partner reported higher levels of perceived control at home. The GLM results supported this association. The two groups of respondents without a partner, produced significantly smaller mean scores of this family control variable as compared to the respondents with a partner. Hypothesis (H2a) was therefore not supported and indeed, it appears that the hypothesised relationship works in the opposing way: the presence of a partner contributes to higher levels of perceived family control. These results contradict recent findings such as those of Rosenbaum and Cohen (1999), who found that self-reliance, as opposed to dependency on a partner, was an important distinction in the estimation of family-based control within their sample of Israeli mothers. Self-reliance in turn, was identified as a significant predictor of well-being outcomes.
The contradictory findings produce by the current research require further investigation. For example, it is possible that the presence of a partner serves to increase levels of female confidence (self-esteem) and this confidence may inflate perceived control scores (see for example: Forthofer, Janz, Dodge, & Clark, 2001; Turner, Lloyd, & Roszell, 1999). In this case, any benefits gained by being the solo decision-maker may therefore be lost. However, the presence of a partner was not associated with any increased levels of job-related control (as might be expected if individual confidence levels were raised). Thus the benefits gained by having a partner were not transferable to the workplace in this situation. Further research is required to clarify this association.
Strong support for Hypothesis (H2b) was produced: high levels of (both types of) perceived control were both associated with, and predictive of, positive well-being outcomes. As stated previously, this direct relationship is firmly established within the literature and the present results reflect the typical strength of these relationships (e.g. Mansell & Brough, 2001; Van der Doef and Maes, 1999).
The influence of a partner: support
No association between the presence of a partner and perceptions of family social support was identified within any of the analyses. Hypothesis (H3a) therefore, was clearly not supported: partners were found to have no effect on the perception of family-based social support. This is an unexpected finding that contradicts the conclusions reached by some recent research (e.g. Nielson et al. 2001). One explanation of this result could be in terms of the measurement of family-based social support used here. The current research did not attempt to define from whom the various sources of family support were obtained. Whilst the respondents with a partner would have been likely to have obtained much of their social support from these partners, it is also likely that the respondents without a partner may have obtained equal amounts of social support from other family members. Thus the actual levels of perceived social support cannot be discriminated by respondent category. It is therefore recommended, that in order to test this hypothesis, subsequent research which evaluates the various sources (and quality) of family social support, is required.
Interestingly, the presence of a partner did produce a small negative association with job-related social support. Thus the respondents with a partner apparently experience (perceive) less social support at work. The GLM analysis reflected these findings in that, lower means scores for this variable, were obtained for both groups of respondents with partners. However, in comparison to the other respondents, these GLM differences were not statistically significant. These results appear to be suggesting that individuals with a partner appear to rely on (rate) social support from these partners to a greater extent, than the support obtained from their work colleagues. Whilst further confirmation of these results is required, this finding does reflect similar suggestions within the literature. For example, it has been observed that social support is obtained more often from partners and family members than from work colleagues and/or formal organisational support facilities (e.g. Brough, 2001; Brough 8,: Mansell, 2001).
Hypothesis (H3b) was generally not supported by this research. Whilst family social support was found to be negatively associated with psychological strain, neither type of social support was identified as a significant predictor within the multiple regression equation. Perceived social support therefore, was not found to be generally associated with psychological well-being. This finding does not reflect other research in this area. Social support has been found to positively influence psychological outcomes in a large number of reported studies (see O'Driscoll & Brough, in press, for an overview). This paper evaluated only the direct effects of social support, it is suggested that further analyses examining the indirect relationships may prove fruitful and may for example, identify the more subtle `third variable' role social support exhibits within the psychological well-being process (e.g. Van der Doef & Maes, 1999).
This paper has demonstrated the influence dependents and partners have upon inter-domain conflict and associated variables. The presence of dependents was found to be strongly associated with increased levels of inter-domain conflict, supporting the conclusions of other studies in this area: Greenhaus & Beutell, (1985); Rosenbaum & Cohen, (1999); Thompson et al. (1999). The experience of inter-domain conflict was especially severe for the respondents with dependents and no partners (solo carers). This finding also reflects other investigations of the psychological well-being of single parents (e.g. Baker & North, 1999). The presence of dependents was also found to be predictive of psychological strain. Thus dependents (children) act to generally decrease levels of well-being. This paper also suggested that the respondents with a partner appear to have improved psychological well-being outcomes, however the presence of a partner per se appeared to have little effect upon levels of inter-domain conflict.
This paper produced a number of unexpected results in regards to associations between the presence of a partner and perceived control and social support. Family control was found to be positively, rather than negatively, associated with the presence of a partner, whilst family social support was unaffected by marital status. Whilst these findings can be partly explained by measurement issues as was described above, the finding for the negative association with control is unexpected and requires subsequent examination. Finally, whilst both types of domain specific control produced a strong relationship with psychological strain as was hypothesised, no such association was identified for the social support predictors. This finding is also unexpected and requires further investigation, particularly in terms of identifying any indirect effects of these domain specific social support measures.
This paper has succeeded in investigating the levels of inter-domain conflict and associated variables, upon different demographic groups of working women in New Zealand. This paper has therefore, contributed to the current interest in inter-domain conflict and psychological well-being. It is suggested here that due to the expected increasing number of untraditional family demographic groups, future research following a similar methodology would also be fruitful, particularly in regard to investigating in more detail the effects of being a lone carer. The current emphasis for New Zealand (and other international) organisations to become more `family friendly' also provides a valuable application of this and future research.
Table 1. Scale descriptives and correlation matrix 1. 2. 3. 4. 1 Partner -- 2 Dependents .10 * -- 3 WFC .02 -.15 ** (.91) 4 FWC .01 -.23 *** .42 *** (.85) 5 Control (family) .15 ** .04 -.10 -.10 6 Control (job) -.04 .05 -.21 *** .07 7 Support (family) -.00 .11 * -.10 * -.20 *** 8 Support (job) .11 .05 -.08 -.01 9 Strain .11 * .05 .27 *** .15 *** 5. 6. 7. 8. 1 Partner 2 Dependents 3 WFC 4 FWC 5 Control (family) (.77) 6 Control (job) .21 *** (.77) 7 Support (family) .36 *** .03 (.92) 8 Support (job) .06 .16 ** .38 *** (.90) 9 Strain -.41 *** -.48 *** -.19 *** -.09 9. Mean SD 1 Partner -- -- 2 Dependents -- -- 3 WFC 20.44 8.45 4 FWC 12.43 6.64 5 Control (family) 35.78 8.33 6 Control (job) 33.47 8.35 7 Support (family) 16.43 5.18 8 Support (job) 14.48 4.68 9 Strain (.92) 17.14 9.47 N.B. WFC = Work [right arrow] Family Conflict; FWC = Family [right arrow] Work Conflict. All tests are two-tailed. Cronbach's alpha coefficients are depicted in parentheses along the diagonal. * p<.05; ** p<.01; *** p<.001. Partner and dependents were both coded as 0 = No, 1 = Yes. Table 2. GLM analysis of respondent categories by job and family conflict, control and social support, and psychological strain Dependent Type III Mean F [[eta].sup.2] variable sum of squares square FWC 754.94 251.65 6.02 *** .01 WFC 623.35 207.78 2.94 * .04 Family social support 141.90 47.30 1.79 .03 Job social support 153.96 51.32 2.41 .03 Family control 922.80 307.60 4.45 ** .05 Job control 190.99 63.66 .89 .02 Psychological strain 771.85 257.28 2.86 * .02 N.B. WFC = Work [right arrow] Family Conflict; FWC = Family [right arrow] Work Conflict. * p<.05; ** p<.01; *** p<.001 Table 3. Means and standard deviations of status for job and family conflict, family control and psychological strain Variable Status M SD N FWC Single 11.70 5.49 46 Solo 14.39 * 8.49 28 Family 13.65 ** 7.00 131 Partner 10.81 5.63 122 WFC Single 19.15 9.02 46 Solo 24.00 * 6.90 28 Family 20.98 8.28 131 Partner 19.31 8.61 122 Family control Single 33.11 * 8.76 46 Solo 32.32 * 7.79 28 Family 36.01 8.18 131 Partner 37.22 8.39 122 Psychological Single 20.44 * 11.45 46 Strain Solo 18.71 8.27 28 Family 16.45 10.21 131 Partner 16.06 8.06 122 N.B. WFC = Work [right arrow] Family Conflict; FWC = Family [right arrow] Work Conflict. * p<.05; *** p<.001 Table 4. Hierarchical multiple regression on psychological strain Step 1 Step 2 Step 3 [R.sup.2] Predictors [beta] [beta] [beta] [R.sup.2] [DELTA] Partner .14 ** .06 .05 .02 * .02 Dependents .05 .11 * .10 * WFC .25 *** .13 ** .23 *** .21 *** Social support -.06 .09 (family) Control (family) -.35 *** -.25 *** FWC .11 * .37 *** .14 *** Social support (job) .04 Control (job) -.40 *** N.B. WFC = Work [right arrow] Family Conflict FWC = Family [right arrow] Work Conflict [beta] = standarised beta coefficients * p<.05; ** p<.01; *** p<.001
The data reported in this paper were collected as part of a larger project, entitled Transactions: Family, Community and Involvement, which was funded by the Foundation for Research, Science and Technology. The financial support of the Foundation for this research is gratefully acknowledged.
We would like to thank the Editor, Mike O'Driscoll for his valuable comments on an earlier draft of this paper.
Amato, P.R. (2000). Diversity within single-parent families. In: D.H. Demo, K.R. Allen, M.A. Fine, (Eds.) Handbook of Family Diversity. Oxford University Press. New York.
Aryee, S., & Luk, V. (1996). Work and non-work influences on the career satisfaction of dual earner couples. Journal of Vocational Behaviour, 49, 38-52.
Aryee, S., Luk, V., Leung, A., & Lo, S. (1999). Role stressors, inter-role conflict, and well-being: The moderating influence of spousal support and coping behaviours among employed parents in Hong Kong. Journal of Vocational Behaviour, 54, 259-278.
Baker, D., North, K. (1999). Does employment improve the health of lone mothers? Social Science and Medicine, 49, 1, 121-131.
Banks, M. H., Clegg, C. W., Jackson, P. R., Kemp, N. J., Stafford, E. M., & Wall, T. D. (1980). The use of the General Health Questionnaire as an indicator of mental health in occupational studies. Journal of Occupational Psychology, 53, 187-194.
Baron, R.M. & Kenny, D.A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology, 51, 6, 1173-1182.
Brough, P.A. (1997). A longitudinal investigation of police officers' psychological well-being. Unpublished PhD Thesis, Bristol University, U.K.
Brough, P. (2001). Experiencing work trauma and work hassles: Identifying best working practices. Proceedings of the British Psychological Society Centenary Annual Conference, March, Glasgow, UK.
Brough, P. & Mansell, A. (2001). The occupational well-being of New Zealand Emergency Service workers: Final report. Unpublished report: School of Psychology, Victoria University of Wellington, New Zealand.
Brown, S.L. (2000). The effect of union type on psychological well-being: Depression among cohabitors versus marrieds. Journal of Health & Social Behavior, 41, 3, 241-255.
Burke, R.J. & Greenglass, E.R. (2001). Hospital restructuring, work-family conflict and psychological burnout among nursing staff. Psychology and Health, 16, 5, 583-594.
Burke, R.J. & McKeen, C.A. (1992). Women in management. In: C.L. Cooper, & I.T. Robertson, (Eds). International Review of Industrial and Organisational Psychology. Wiley, Chichester.
Carlson, D. S., Kacmar, K. M., & Williams, L. J. (2000). Construction and initial validation of a multidimensional measure of work-family conflict. Journal of Vocational Behaviour, 56, 249-276.
Coombs, R.H. (1991). Marital status and personal well-being: A literature review. Family Relations: Journal of Applied Family & Child Studies, 40, 1, 97-102.
Eagle, B. W., Miles, E. W., & Icenogle, M. L. (1997). Inter-role conflicts and the permeability of work and family domains: Are there gender differences? Journal of Vocational Behaviour, 50, 168-184.
Forthofer, M.S., Janz, N.K., Dodge, J.A. & Clark, N.M. (2001). Gender differences in the associations self esteem, stress and social support with functional health status among older adults with heart disease. Journal of Women and Aging, 13, 1, 19-36.
Frone, M., Russell, M., & Cooper, C. L. (1992). Antecedents and outcomes of work-family conflict: Testing a model of the work-family interface. Journal of Applied Psychology, 77, 65-78.
Gjerdingen, D., McGovern, P., Bekker, M., Lundberg, U. & Willemsen, T. (2000). Women's work roles and their impact on health, well-being and career: Comparisons between the United States, Sweden and the Netherlands. Women & Health, 31, 4, 1-20.
Goldberg, D.P. (1972). The Detection of Psychiatric Illness by Questionnaire. Oxford University Press, Oxford.
Greenhaus, J. H., & Beutell, N. J. (1985). Sources of conflict between work and family roles. Academy of Management Review, 10, 1, 76-88.
Hammer, L. B., Allen, E. & Grigsby, T. D. (1997). Work-family conflict in dual-earner couples: Within-individual and crossover effects of work and family. Journal of Vocational Behaviour, 50, 185-203.
Herman, J.B. & Gyllstrom, K.K. (1977). Working men and women: Inter- and intra-role conflict. Psychology of Women Quarterly, 1, 4, 319-333.
Hochschild, A. (1989). The Second Shift. New York: Avon Books.
Karasek, R. A. (1979). Job demands, job decision latitude and mental strain: Implications for job redesign. Administrative Science Quarterly, 24, 285-308.
Kline, M., & Cowan, P. A. (1988). Re-thinking the connections among "work" and "family" and well-being. Journal of Social Behaviour and Personality, 3, 61-90.
Lazarus R. S., & Folkman, S. (1984). Coping and adaptation. In W. D. Gentry (Ed.), The Handbook of Behavioural Medicine. New York: Guilford.
Mansell, A. & Brough, P. (2001). Occupational support and control in New Zealand Custom Officers. Proceedings of the New Zealand Psychological Society Conference. University of Auckland, Auckland, August.
Netemeyer, R. G., Boles, J. S., & McMurrian, R. (1996). Development and validation of work-family conflict and family-work conflict scales. Journal of Applied Psychology, 81, 4, 400-410.
Nielson, T.R., Carlson, D.S., Lankau, M.J, (2001). The supportive mentor as a means of reducing work-family conflict. Journal of Vocational Behavior, 59, 3, 364-381.
O'Driscoll, M. P. (1996). The interface between job and off-job roles: Enhancement and conflict. International Review of Industrial and Organisational Psychology, 11, 279-306.
O'Driscoll, M. (2000). Work and family transactions. In: P.Koopman-Boyden et al. (Eds.) Transactions in the mid-life family (pp.92-112). Hamilton: Population Association of New Zealand.
O'Driscoll, M. &. Brough, P.A. Job stress and burnout. In: M. O'Driscoll, P. Taylor & T. Kalliath, (Eds). Organizational Psychology in Australia and New Zealand. Oxford University Press (in press).
O'Driscoll, M. P., & Humphries, M. (1994). Time demands, interrole conflict and coping strategies among managerial women. International Journal of Employment Studies, 2, 57-75.
Oliver, J. & Brough P. Primary appraisal, negative affectivity and psychological well-being. New Zealand Journal of Psychology, (in press).
Parasuraman, S., Greenhaus, J. H., & Granrose, C. S. (1992). Role stressors, social support, and well-being among two-career couples. Journal of Organisational Behaviour, 13, 339-356.
Parasuraman, S., Purohit, Y. S., & Godshalk, V. M. (1996). Work and family variables, entrepreneurial career success, and psychological well-being. Journal of Vocational Behaviour, 48, 275-300.
Parker, S.K., Wall, T.D. & Myers, C. (1995). The effects of a manufacturing initiative on employee jobs and strain. In S.A. Robertson, Contemporary Ergonomics. Taylor & Francis: London.
Pearlin, L. I., & Schooler, C. (1978). The structure of coping. Journal of Health and Social Behaviour, 19, 2-21.
Pleck, J. H., Staines, G. L., & Lang, L. (1980). Conflicts between work and family life. Monthly Labour Review, 103, 3, 29-32.
Richardsen, A.M., Burke, R.J., Mikkelsen, A. (1999). Job pressures, organizational support, and health among Norwegian women managers. International Journal of Stress Management, 6, 3, 167-177.
Rosenbaum, M., & Cohen, E. (1999). Equalitarian marriage, spousal support, resourcefulness, and psychological distress among Israeli working women. Journal of Vocational Behaviour, 54, 102-113.
Rudman, R. (1999). Human Resources Management in New Zealand. (3rd edition.) Auckland: Longman.
Skues, J. S., & Kirkby, R. J. (1995). Women in the workforce: Gender differences in occupational stress and coping. In: P. Cotton (Ed). Psychological health in the workplace: Understanding & managing occupational stress. Carlton South VIC, Australia: Australian Psychological Society Ltd.
Statistics New Zealand (2001). Key Statistics: April 2001. Wellington: Government Press.
Tabachnick, B.G. & Fidell, L.S. (1996). Using Multivariate Statistics. Harper Collins, New York.
Thoits, P.A. (1987). Gender & marital status differences in control & distress: Common stress verses unique stress explanations. Journal of Health & Social Behaviour, 28, 7-22.
Thompson, C. A., Beauvais, L. L., & Lyness, K. S. (1999). When work-family benefits are not enough: The influence of work-family culture on benefit utilisation, organisational attachment, and work-family conflict. Journal of Vocational Behaviour, 54, 392-415.
Turner, R.J., Lloyd, D.A. & Roszell, P. (1999). Personal resources and the social distribution of depression. American Journal of Community Psychology, 2 7, 5, 643-672
Van der Doef, M., & Maes, S. (1999). The job demand-control(-support) model and psychological well-being: A review of 20 years of empirical research. Work and Stress, 13, 2, 87-114.
Voydanoff, P. (2002). Linkages between the work-family interface and work, family, and individual outcomes: An integrative model. Journal of Family Issues, 23, 1, 138-164.
Zapf, D., Dormann, C. & Frese, M. (1996) Longitudinal studies in organisational stress research: A review of the literature with reference to methodological issues. Journal of Occupational Health Psychology, 1, 2, 145-169.
Address for correspondence: Dr Paula Brough School of Applied Psychology Mt Gravatt Campus, Griffith University Brisbane, Queensland 4111 Australia. Telephone: +61 7 3875 3378 Facsimile: +61 7 3875 3388 Email: P.Brough@mailbox.gu.edu.au Paula Brough School of Applied Psychology, Griffith University, Brisbane Anouk Kelling School of Psychology, Victoria University of Wellington
|Printer friendly Cite/link Email Feedback|
|Author:||Brough, Paula; Kelling, Anouk|
|Publication:||New Zealand Journal of Psychology|
|Date:||Jun 1, 2002|
|Previous Article:||Effects of iron treatment on cognitive performance and working memory in non-anaemic, iron-deficient girls.|
|Next Article:||Psychometric properties of three scales of depression and well-being in a mature New Zealand sample.|