This article was downloaded by: [New York University] On: 08 January 2015, At: 12:25 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Women & Health Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/wwah20
Worker and Mother Roles, Spillover Effects, and Psychological Distress a
Rosalind C. Barnett PhD & Nancy L. Marshall EdD
a
a
Center for Research on Women, Wellesley College, Wellesley, MA Published online: 05 Nov 2010.
To cite this article: Rosalind C. Barnett PhD & Nancy L. Marshall EdD (1992) Worker and Mother Roles, Spillover Effects, and Psychological Distress, Women & Health, 18:2, 9-40, DOI: 10.1300/J013v18n02_02 To link to this article: http://dx.doi.org/10.1300/J013v18n02_02
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
Downloaded by [New York University] at 12:25 08 January 2015
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/ page/terms-and-conditions
Worker and Mother Roles, Spillover Effects, and Psychological Distress
Downloaded by [New York University] at 12:25 08 January 2015
Rosalind C. Barnett, P h D Nancy L. Marshall, EdD
ABSTRACT. This paper exatnines the relationship between the occupancy and quality of multiple-roles and psychological distress in a stralified random sample of 403 women employed as licensed practical nurses and social workers. We examined the main effects of the quality of the employee and parent roles and the interaction effects between these variables. Negative- and positive-spillover effects, frotn job to parenting and from parenting to job, were examined in an attempt to illuminate the processes by which . multiple roles affect employed mothers' vulnerability or resilience to psychological distress. We found no negative-spillover effects from job to parenting or from parenting to job, but we did find positive-spillover effects frotn job to parenting. Women with rewarding jobs were protected from the negative mental-health effects of troubled relationships with their children. This protection a c c ~ e dlo employed mothers regardless of lheir partnership stalus or the age of their children. Although based on cross-sectional Rosalind C. Barnett and Nancy L. Marshall are affiliated wilh the Center for Research on Women at Wellesley College, Wellesley, MA. This paper is based on data from a short-tm longitudinal study of workplace and non-workplace stressors and mitigators of stress among women health-care providers. Data for this paper come from the first of three interviews and were collected in 1986-87. This project was funded by the National Inslitute for Occupalional Safety and Health (# OH0 1968). We are indebted to Judith Singer for her careful readings and critiques of earlier versions. O 1992
Women & Health. Vol. 18(2) 1992 by The Haworth Press, Inc. All rights reserved.
9
10
WOMEN & HEALTI1
Downloaded by [New York University] at 12:25 08 January 2015
analyses, these findings suggest mechanisms by which employed mothers reap a mental-health advantage from multiple roles, even when some of Ulose roles are stressful. The relationship between psychological distress and social roles is receiving renewed interest in the wake of the large numbers of employed women who are mothers. Are employed women with children at greater risk for psychological distress than are employed women without children? Is the relationship between distress and experiences in one role affected by experiences in the other role? Are the effects independent or interactive? Although these questions have important theoretical implications and are at the heart of much heated discussion, surprisingly little research has addressed them specifically. Despite general concern about the impact of multiple roles on women's mental health, most studies examine only the impact of individual roles.' Research on mental health has generally addressed the issue of the relative mental health of employed versus nonemployed women and of mothers versus non-mothers. We know, for example, that employed women report lower levels of anxiety and depression than non-employed women (Barnett & Baruch, 1985; Brown &Harris, 1978; Thoits, 1983), and that mothers report more symptoms of distress than non-mothers (Bamett & Baruch, 1985; McLanahan & Adams, 1987; Veroff, Kulka & Douvan, 1981). This pattern is strongest among mothers with children living at home under eighteen years of age, especially among those with preschool-age children (Pearlin, 1975; Umberson & Gove, 1989). Recent work on multiple roles, which indicates that the more roles a person (male or female) occupies the better their mental health (Sieber, 1974; Thoits, 1983), suggests that employed mothers would report fewer symptoms of psychological distress than employed non-mothers. However, considerations of role conflict and role strain suggest that combining the very demanding roles of mother and paid employee would be particularly stressful and might, therefore, produce more symptoms of psychological distress. Many researchers caution that to understand the relationship between multiple roles and mental health, a focus on role occupan-
Downloaded by [New York University] at 12:25 08 January 2015
Rosalind C. Borncll and Nancy L. Marslrall
I1
cy per se is less fruitful than a focus on role quality (Barnett & Baruch, 1985; Kotler & Wingard, 1989). Under this theory, although occupancy of the role of mother may not, predict distress levels among employed women, the quality of their experience in the maternal role might. One mechanism or linkage posited for the hypothesized negative impact of multiple roles on psychological distress is that of "spillover." It has been suggested that spillover from family to work is more likely among employed women, whereas spillover from work to family is more common among men (Pleck, 1977). Accordingly, employed women with family roles would bring their home-based woes to the job, thereby rendering them less able to cope with the demands of the job, and by extension, more vulnerable than women without family roles (and men) to psychological distress. This thinking suggests greater psychological distress associated with negative-spillover effects from family to job for women. The other version of negative-spillover effects, i.e., from work to family, has generated much of the concern expressed in the popular media about the disastrous effects of work on the children and husbands of employed women (Schwartz, 1989). In this version, women bring home their work-related problems, which then contaminate their home life, resulting in heightened distress for themselves as well as their children. In contrast, the notion of positive-spillover effects, i.e., that distress would be lowered by work rewards offsetting parental woes or parental rewards offsetting work woes, has received little attention (see Crouter, 1984 for an exception). One study of women's subjective well-being reported a positive-spillover effect from work to parenting (Barnett & Baruch, 1985). Among employed mothers, having a rewarding job role ameliorated the relationship between problematic mother-child relationships and subjective well-being. Evidence of negative-spillover effects on psychological distress would support the role-strain and role-conflict models underlying the relationship between stress and illness. Evidence of positivespillover effects would suggest a mechanism by which the mental health of employed mothers would benefit from multiple roles, as long as the moderating role was rewarding.
WOMEN & HEALTH
Downloaded by [New York University] at 12:25 08 January 2015
OPERATIONALIZING SPILLOVER How are we to operationalize spillover? As a first step, we consider two models of the relationship between roles and psychological distress, following the work of Wheaton (1982), House (1981), and Cohen and Wills (1985): one model in which the quality of experience in one role does not affect the relationship between psychological distress and the quality of experience in another role, and another model in which it does. These models-a main-effects model and an interaction-effects model-are shown in Figures la and Ib. Although the arrows imply that role quality affects distress, it is equally possible that distress affects role quality. Women who have high levels of psychological distress may experience greater role-related concerns and fewer role-related rewards. With cross-sectional data, as in this paper, it is not possible to determine the direction of causation. Under the main-effects model, the quality of experience in one role does not affect the relationship between the quality of experience in the other role and psychological distress. With respect to the effects of the employee and mother roles, the main-effects model indicates no compounding of effects from one role to the other. To illustrate, the home-based woes or joys of employed mothers would not compound the effects of their work experience on their level of psychological distress. Most researchers studying the effects of role quality on mental-health states examine only one role at a time, implicitly assuming this main-effects model. For example, studies of the relationship between work stress and health outcomes often do not assess family-role quality (see for example, Karasek, Schwartz, Theorell, Pieper, Russell, & Michela, 1982). In contrast, the interaction-effects model posits that emotional feelings generated in one arena-work or family-"spillover" into the other, increasing vulnerability or resilience to psychological distress. To demonstrate spillover effects on psychological distress, one has to show that the interaction between the experiences in the two roles accounts for a significant proportion of the variance in psychological distress, over and above that accounted for by the main effects of the experiences in the two roles. This approach
Downloaded by [New York University] at 12:25 08 January 2015
FIGURE la. Main-effects model of the relationship between job-role and parent-role quality and psychological distress.
WORK-ROLE QUALITY
PSYCHOLOGlCAL DISTRESS PARENT-ROLE QUALITY
c
Downloaded by [New York University] at 12:25 08 January 2015
Downloaded by [New York University] at 12:25 08 January 2015
Rosalind C . Barr~cttand Nancy L. Marslull
15
differs from that taken by most research on role quality across domains, which does not look at health outcomes. The typical study documents correlational relationships between affectiveexperiences in one role and affective experiences in another role. Moreover, most of these studies have been done on male samples (Evans & Bartolome, 1980, 1984) and report positive correlations between occupational stress and family dissatisfaction. In a study of men and women, Crosby (1984) reported a positive correlation between job satisfaction and home satisfaction. Such findings do not address the question of whether these correlational relationships have mentalhealth consequences. While an interaction between the quality of experiences in two roles will test the existence of spillover, we would also like to disaggregate negative spillover from positive spillover. To do this, we use two different measures, one of the positive aspects, or rewards, of a role, and the other of the negative aspects, or concerns, of a role. We then examine whether either the rewards of Role A, or the concems of Role A, interact with the overall quality of Role B. If there is an interaction between the rewards of Role A and the overall quality of Role B, we can examine a graph of this relationship to determine whether there is positive spillover from Role A to Role B, that is, whether women with rewards in Role A are at reduced risk for psychological distress compared to other women with similar experiences in Role B. Similarly, if there is an interaction between the concems of Role A and the overall quality of Role B, we can examine a graph of this relationship to determine whether there is negative spillover from Role A to Role B. We can also test for positive and negative spillover from Role B to Role A by examining whether the rewards or the concerns of Role B interact with the overaN quality of Role A. Using data from a disproportionately-stratified random sample of 403 employed mothers, ages 25 to 55, this study addresses the following questions: 1. Are employed mothers more distressed than working women without children? 2. Do women's experiences at work and as parents have main effects on their experience of psychological distress'?
WOMEN & NEALTti
16
3. Is there positive or negative spillover from one role to another that contributes to psychological distress?
METHODS
Downloaded by [New York University] at 12:25 08 January 2015
Sample The sample for this paper consists of 403 employed women, who were part of a larger study of employed women. The data come from the first wave of a three-wave longitudinal study (over two years) of a disproportionately-stratified random sample of 403 Massachusetts women employed at least half-time in one of two health-care professions-licensed practical nursing and social work. We selected these two professions because (1) they are predominantly female. (2) they produce high job strain? and (3) complete public licensure records are available. Since the Commonwealth of Massachusetts requires registration as a condition of practicing these professions, the professional registries comprise the entire population of registered social workers and LPNs in the state. Within each occupation, the sample was stratified by race (White or Black), parental status (has a child or does not have a child) and partnership status (marriedfliving with a partner or not being in either type of relationship). Because of the low percentage of Black employed women listed in the registries (e.g., only 4% of registered social workers were Black and employed), we oversampled Blacks to obtain a group large enough for analysis. In addition, because the role-status variables are confounded naturally in the population (e.g.. Black women 25-55 are less likely than White women 25-55 to be currently married), we further stratified the sample within occupation by race, partnership status, and parental status. The success of our stratification is reflected in the low correlations among occupation, race, partnership status and parental status. For example, among social workers, the correlation between race and partnership status is .11; between race and parental status, the correlation is -.11. The comparable figures for the LPNs are .04 and -.06.
Downloaded by [New York University] at 12:25 08 January 2015
Rosalind C. Borne11 and Nancy L. Marsl~all
17
Of the potential subjects whom we contacted, only 4% refused to participate. The total sample consisted of 403 women; 248 social workers (61.5%) and 155 licensed practical nurses (38.5%). Sixtyone women (15.3%) were Black, and 342 (84.7%) were White. The mean age of the respondents was 39.5 years (sd = 7.4). Of the 198 women who were partnered (49.1%). 75 did not have children, 123 did. Of the 205 women (50.9%) who were single, 99 did not have children, 106 did. Most of the mothers' (72%) youngest child is school-aged, that is, between the ages of 6 and 18 years; 13.9% had a child under age six, only 14% of the mothers do not have a child under the age of 19. Finally, although the average number of children was 2.5, the mean number of children currently living at home was 1.5 (sd = 1.2). On average, the respondents had been working in their respective fields for 11 years (the range was 2 to 35 years) and at their current jobs for 5.7 years (sd = 5.1). They worked 39.4 hours per week, on average (sd = 8.7). and 80% worked the same schedule on a regular basis. The mean individual income in 1985 was $24,400 (sd = $2,700). All resoondents lived within a 25-mile radius of Boston. Respondents were interviewed in their homes or offices by a trained interviewer. Interview questions were structured and responses were recorded by the interviewer, using pre-coded response categories; Interviews lasted about 2 hours and covered the quality of the woman's experiences in each of her major social roles, i.e.. employee, mother, partner, as well as indices of psychological distress, well-being, and physical health. Respondents were paid $10 for participating. Although the sample is representative of Massachusetts women employed in these two heavily female occupations, it is not representative of women employed in a wider range of occupations. It is not clear whether or how predispositions to such jobs affects the findings of the study. One possibility is that women who assume nurturant, caring jobs are less likely than women not in such jobs to compartmentalize their affective experiences. They might be expected to carry problems with their children with them to the workplace and to bring their emotional distresses home with them, increasing their vulnerability to the negative effects of stressors in their other role. Conversely, emotionally rewarding
I8
W O M M & HEALTH
experiences in one role might have special meaning to these women. thereby increasing their resilience to the effects of stressors in their other role. As a result, these women might be more likely than women not attracted to these fields to experience spillover effects, negative as well as positive, both from job to home and from home to job.
Downloaded by [New York University] at 12:25 08 January 2015
Measures Psychological distress. Psychological distress was assessed by the anxiety and depression subscales of the SCL-90-R, a frequency of symptoms measure (Derogatis. 1975; 1983). Respondents indicate on a 5-point scale (from 0 = not at all, to 4 = extremely) how often in the past week they were bothered by each of 14 symptoms of anxiety and 10 symptoms of depression. The decision to combine the scales into a psychological distress score was based on the high correlation ( r = .80) between the scales and on the similarity in the pattern of correlations between the anxiety and depression scales and the other variables of interest in the study. The use of combined scale scores follows the work of Barnett and Marshall (1991) and Folkman, Lazarus, Gruen, and DeLongis (1986). The SCL-90-R has high levels of both internal consistency and test-retest reliability. In this sample, coefficient alpha was .88 for depression and .89 for anxiety. These figures are similar to those reported by Derogatis (1983). Satisfactory test-retest correlations (.82 for depression and .80 for anxiety) have also been reported (Derogatis, 1983). Role rewards and role concerns. Positive and negative aspects of the roles of worker and mother were assessed by reward and concern scales constructed originally from data gathered during extensive interviews with 72 women, ages 35 to 55 (Baruch & Bamett, 1986 for a full discussion). These scales have been revised and expanded, drawing on psychometric data from a sample of 238 women, focus groups with an additional 30 women employed as licensed practical nurses and social workers, as well as the work of other researchers. For each role, respondents are asked to think about their situation as it is right now and to indicate on a 4-point scale (I =not at.all, to 4 = extremely) to what extent, if at all, each
Downloaded by [New York University] at 12:25 08 January 2015
Rosalind C . Barnc11 and Nancy L. Marsl~all
19
of the items is rewarding (or of concern).' (The number of items differs across roles; for the role of worker, there are 25-reward and 25-concern items; for the role of parent, there are 18-reward and 20-concern items.) For example, for the role of paid worker, each subject was asked how rewarding she found "the job security" and to what extent was "the job's not using your skills" a concern. For the role of mother, each woman with children was asked how rewarding she found "the love they show" and how'much of a concern was "how they spend their free time" (items comprising the reward and concern scales are shown in Appendix A). Each subject received two scores for each social role: a reward score and a concern score. Test-retest reliability coefficients, calculated on a 10% random subsample retested within 1-3 months of the initial interview, were ..88 for both job rewards and job concerns, .82 for parent rewards, and .70 for parent concerns. Internal consistency reliability coefficients (Cronbach alphas) for job rewards was 3 8 ; for job concerns, it was -89; for parent rewards, .83; and for parent concerns, .89. The job rewards and job concerns scores are correlated -.57. Overall role q~~ality. Overall role quality is conceptualized as the balance between the positive and negative aspects of a role, and is operationalized as the sum of the reward scores and the inverse of the concern scores for each role, adjusted for scale length. This score provides an index of the overall quality of experience women have in their job and parent roles. Previous research indicates that overall role quality is associated with psychological distress (Bamett & Baruch, 1985; Baruch & Barnett, 1986; Bamett, Davidson, & Marshall, 1991; Bamett, Kibria, Marshall, & Pleck, 1991; Bamett & Marshall, 1991). RESULTS
Preliminary analyses comparing the two occupational groups indicated no significant differences in the means for either the jobreward, job-concern, parent-reward, or parent-concern scales. Using a dummy variable for occupation, a series of regression models was
Downloaded by [New York University] at 12:25 08 January 2015
20
WOMEN & HEALTH
estimated to test for the main effects of occupation on psychological distress, as well as interactions between occupation and race, age, and per capita income. Because occupation was never a statistically significant predictor, nor did the main effect of role quality differ by occupation, the two occupational groups were combined for the analyses reported in this paper. Correlations between the reward and concern scores for the job and parent roles are shown in Table 1. As noted above, within each role, rewards and concerns are negatively correlated. Employed women who report more positive experiences at work, report fewer negative experiences at work; similarly, employed mothers who report morepositive experiences in their maternal role, report fewer concerns with their children. These moderate negative correlations between the rewarding and problematic aspects of these roles are expected, because "'good' jobs tend to be high on rewards and low on concems while 'bad' jobs tend to have problems in more than one area" (Hackman and Oldham, 1975, p. 166). ,In fact, the magnitude of the correlations between the job rewards and job concerns scores indicate that the job rewards score explains only 32% of the variance in the job concerns score; fully 68% of the variance in the job concerns score is unique to that measure. Similarly, parent rewards and parent concerns are correlated -.22; 95% of the variance in parent concems is unique to that measure. As expected, the correlations between job-role quality and psychological distress ( r = -.42, p < .001) and between parent-role quality and psychological distress (r = -.23, p < .001) were negative and significant. Employed women who report more positive experiences at work, report fewer symptoms of psychological distress. Employed mothers who report more positive experiences in their parenting role report lower levels of psychological distress. The correlations between the reward and concern scores in each role and psychological distress range between -.07, ns and .37, p < .001 and are in the expected direction (i.e., job and parent rewards are correlated negatively with distress; job and parent concerns are correlated positively). Role concern account scores for approximately 12.5% of the variance 'in psychological distress, whereas role reward scores account for between -50% and 10.0%.
21
Rosalind C. Darrieff and Namy L. Marslull
TABLE 1. In~ercorrelationsBctween Rewvd and Concern Scores for Job and Pnrenl Rolcs
Downloaded by [New York University] at 12:25 08 January 2015
Job Rewards Job Qncerns Parent Rewards parent ~ o ~ m r n s
Note. *p
-.39***
.17** -.07
.14*
-.17* .36***
-.31*** -.22**
= 211
< .05; **p < .Ol; ***g < .001.
Research Question 1: Are Employed Mothers More Distressed Than Employe'd Women Without Children? To examine the relationship between parental status and psychological distress, a dummy variable for parental status ( I = parent, 0 = non-parent) was entered as a predictor into a regression model that included a set of four controls: age, race, education, and per capita income. The sample for this analysis consisted of all 403 employed women. Although the overall regression model was significant, parental status was not. This finding indicates that among employed women, parental status per se is not a significant predictor of psychological distress; in other words, after controlling for age, race, education and per capita income, employed mothers
Downloaded by [New York University] at 12:25 08 January 2015
22
WOMEN & HEALTH
are at no greater risk of psychological distress than are employed women without children. Other factors. There are data to suggest that single mothers are at higher risk for stress-related mental-health problems than are partnered mothers (McLanahan & Adams, 1987). Firm conclusions about the reasons for the vulnerability of single mothers are questioned by recent findings indicating that after controlling statistically for the effects of per capita income, occupation, and education, employed single mothers are at no greater risk for psychological distress than are employed partnered mothers (Bamett & Marshall, 1991). Risk for depression among mothers has also been related to the number of preschool-age children at home (McLanahan & Adams, 1987; Pearlin, 1975). Thus, although the main focus in this paper is on the roles of paid employee and mother, in every analysis the effects of partnership status, number of children living at home, number of preschool children, and having children eighteen years of age or younger are also examined. In this instance, we first tested the main and interactive effects of partnership status in predicting psychological distress, trying to ascertain whether employed single mothers report higher levels of distress than employed partnered mothers. We examined two models: one with the main effects of partnership status (I = partnered, 0 = single) and another with the main effect of partnership status and the interaction between parental status and partnership status. Neither the main nor interaction effects was significant. Thus, among employed women, after considering the effects of age, race, education, and per capita income, psychological distress is unrelated to parental status, regardless of whether they are partnered or single. After examining separately the main effects of number of children at home, number of preschool children, and having children eighteen years of age or younger, we examined interactions between each of the predictors and partnership status by creating the appropriate interaction terms, (e.g., partnership status x number of preschool children, partnership status x number of children living at home). Once again, all effects were non-significant. Thus, among employed mothers, after considering the effects of age, race,
Rosalind C. Borne11 and Nancy L. Marshall
23
education and per capita income, psychological distress is unrelated to parental status, independent of whether the employed mothers were partnered or single, or had preschool-age children, had children eighteen years of age or younger, or had few or many children living at home.
Downloaded by [New York University] at 12:25 08 January 2015
Research Question 2: Do Women's Experiences at Work and as Parents Have Main Effects on Their Experience of Psychological Distress? Analyses of the relationship between psychological distress and parent-role quality and job-role quality were conducted on the 228 employed mothers4 in the sample who had complete demographic data. Not surprisingly, the subsample of employed mothers differed significantly from the total sample with respect to age and per capita'income (Table 2). The mothers were significantly older and, as a consequence of their larger family size, had lower per capita income. Employed mothers were also more likely than employed non-mothers to be LPNs than social workers, however, they did not differ on race. Fifty and nine-tenths percent (n = 116) of the employed mothers were married, 36.4% (n = 83) were divorced, 12.7% (n = 29) were never-married or separated. Not surprisingly, parent-role quality shows the predicted relationship with distress, even after controlling for the effects of mothers' age, race, education, per capita income, and job-role quality (Table 3). Similarly, job-role quality is significantly related to psychological distress, .even after controlling for age, race, education, per capita income and parent-role quality. The regression equation was significant, (F (6, 204) = 4.77, p < .001, RZ = .12. While psychological distress is unrelated to parental status, employed mothers with positive mother-child experiences report lower levels of psychological distress, and employed mothers with troubled mother-child relationships report higher levels. Thus, the quality of the maternal role, not the occupancy of the role, predicts distress among employed mothers.
WOMEN & HEALTH
24
Downloaded by [New York University] at 12:25 08 January 2015
TABLE 2. Descriptive Data Comparing Employed Mothers to Employed Nonmothers
42.7***
6.43
35.5
15,559***
9301
27,397
ompatima
1.56**
.50
1.69
.46
~aceb
1.11
.32
1.18
.30
Age
Per Capita Imome
6.61
U,324
a 1 = Licaszd Fractical Nurse; 2 = Social Wcnksr b 1 = Wni-;
2 = Black
Research Question 3: Is There Positive or Negative Spillover from One Role to Another That Contributes to Psychological Distress? Consistent with previous research, we found a significant positive correlation ( r = .29, p < .001)between parent-role quality and jobrole quality. Employed mothers who report higher levels of positive-role quality in one role tend to report higher levels of positive-role quality in their other role. However, any one role explains less than 10% of the variance in the other role. We then examined whether these associations between the rolequality measures translate into spillover effects. operationalized as the interaction between the measures in predicting psychological
Rosalind C. Bartiell and Nancy L. Morslwll
25
TABLE 3. ~ a i n - ~ f f i cModel: ts Relationship of Parent-Role and Job-Role Quality to Psychological Distress
we
Downloaded by [New York University] at 12:25 08 January 2015
Race
EcWation Per Capita IKIlmne
Parent-Role Qality Job-Role puality
Note.
= 211
a Unstmlardlzed regression xefficients b Startiard error
of the unstandardized regression coefficient
c Standardized regression caeffioents p e0.05; *%
Women & Health Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/wwah20
Worker and Mother Roles, Spillover Effects, and Psychological Distress a
Rosalind C. Barnett PhD & Nancy L. Marshall EdD
a
a
Center for Research on Women, Wellesley College, Wellesley, MA Published online: 05 Nov 2010.
To cite this article: Rosalind C. Barnett PhD & Nancy L. Marshall EdD (1992) Worker and Mother Roles, Spillover Effects, and Psychological Distress, Women & Health, 18:2, 9-40, DOI: 10.1300/J013v18n02_02 To link to this article: http://dx.doi.org/10.1300/J013v18n02_02
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
Downloaded by [New York University] at 12:25 08 January 2015
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/ page/terms-and-conditions
Worker and Mother Roles, Spillover Effects, and Psychological Distress
Downloaded by [New York University] at 12:25 08 January 2015
Rosalind C. Barnett, P h D Nancy L. Marshall, EdD
ABSTRACT. This paper exatnines the relationship between the occupancy and quality of multiple-roles and psychological distress in a stralified random sample of 403 women employed as licensed practical nurses and social workers. We examined the main effects of the quality of the employee and parent roles and the interaction effects between these variables. Negative- and positive-spillover effects, frotn job to parenting and from parenting to job, were examined in an attempt to illuminate the processes by which . multiple roles affect employed mothers' vulnerability or resilience to psychological distress. We found no negative-spillover effects from job to parenting or from parenting to job, but we did find positive-spillover effects frotn job to parenting. Women with rewarding jobs were protected from the negative mental-health effects of troubled relationships with their children. This protection a c c ~ e dlo employed mothers regardless of lheir partnership stalus or the age of their children. Although based on cross-sectional Rosalind C. Barnett and Nancy L. Marshall are affiliated wilh the Center for Research on Women at Wellesley College, Wellesley, MA. This paper is based on data from a short-tm longitudinal study of workplace and non-workplace stressors and mitigators of stress among women health-care providers. Data for this paper come from the first of three interviews and were collected in 1986-87. This project was funded by the National Inslitute for Occupalional Safety and Health (# OH0 1968). We are indebted to Judith Singer for her careful readings and critiques of earlier versions. O 1992
Women & Health. Vol. 18(2) 1992 by The Haworth Press, Inc. All rights reserved.
9
10
WOMEN & HEALTI1
Downloaded by [New York University] at 12:25 08 January 2015
analyses, these findings suggest mechanisms by which employed mothers reap a mental-health advantage from multiple roles, even when some of Ulose roles are stressful. The relationship between psychological distress and social roles is receiving renewed interest in the wake of the large numbers of employed women who are mothers. Are employed women with children at greater risk for psychological distress than are employed women without children? Is the relationship between distress and experiences in one role affected by experiences in the other role? Are the effects independent or interactive? Although these questions have important theoretical implications and are at the heart of much heated discussion, surprisingly little research has addressed them specifically. Despite general concern about the impact of multiple roles on women's mental health, most studies examine only the impact of individual roles.' Research on mental health has generally addressed the issue of the relative mental health of employed versus nonemployed women and of mothers versus non-mothers. We know, for example, that employed women report lower levels of anxiety and depression than non-employed women (Barnett & Baruch, 1985; Brown &Harris, 1978; Thoits, 1983), and that mothers report more symptoms of distress than non-mothers (Bamett & Baruch, 1985; McLanahan & Adams, 1987; Veroff, Kulka & Douvan, 1981). This pattern is strongest among mothers with children living at home under eighteen years of age, especially among those with preschool-age children (Pearlin, 1975; Umberson & Gove, 1989). Recent work on multiple roles, which indicates that the more roles a person (male or female) occupies the better their mental health (Sieber, 1974; Thoits, 1983), suggests that employed mothers would report fewer symptoms of psychological distress than employed non-mothers. However, considerations of role conflict and role strain suggest that combining the very demanding roles of mother and paid employee would be particularly stressful and might, therefore, produce more symptoms of psychological distress. Many researchers caution that to understand the relationship between multiple roles and mental health, a focus on role occupan-
Downloaded by [New York University] at 12:25 08 January 2015
Rosalind C. Borncll and Nancy L. Marslrall
I1
cy per se is less fruitful than a focus on role quality (Barnett & Baruch, 1985; Kotler & Wingard, 1989). Under this theory, although occupancy of the role of mother may not, predict distress levels among employed women, the quality of their experience in the maternal role might. One mechanism or linkage posited for the hypothesized negative impact of multiple roles on psychological distress is that of "spillover." It has been suggested that spillover from family to work is more likely among employed women, whereas spillover from work to family is more common among men (Pleck, 1977). Accordingly, employed women with family roles would bring their home-based woes to the job, thereby rendering them less able to cope with the demands of the job, and by extension, more vulnerable than women without family roles (and men) to psychological distress. This thinking suggests greater psychological distress associated with negative-spillover effects from family to job for women. The other version of negative-spillover effects, i.e., from work to family, has generated much of the concern expressed in the popular media about the disastrous effects of work on the children and husbands of employed women (Schwartz, 1989). In this version, women bring home their work-related problems, which then contaminate their home life, resulting in heightened distress for themselves as well as their children. In contrast, the notion of positive-spillover effects, i.e., that distress would be lowered by work rewards offsetting parental woes or parental rewards offsetting work woes, has received little attention (see Crouter, 1984 for an exception). One study of women's subjective well-being reported a positive-spillover effect from work to parenting (Barnett & Baruch, 1985). Among employed mothers, having a rewarding job role ameliorated the relationship between problematic mother-child relationships and subjective well-being. Evidence of negative-spillover effects on psychological distress would support the role-strain and role-conflict models underlying the relationship between stress and illness. Evidence of positivespillover effects would suggest a mechanism by which the mental health of employed mothers would benefit from multiple roles, as long as the moderating role was rewarding.
WOMEN & HEALTH
Downloaded by [New York University] at 12:25 08 January 2015
OPERATIONALIZING SPILLOVER How are we to operationalize spillover? As a first step, we consider two models of the relationship between roles and psychological distress, following the work of Wheaton (1982), House (1981), and Cohen and Wills (1985): one model in which the quality of experience in one role does not affect the relationship between psychological distress and the quality of experience in another role, and another model in which it does. These models-a main-effects model and an interaction-effects model-are shown in Figures la and Ib. Although the arrows imply that role quality affects distress, it is equally possible that distress affects role quality. Women who have high levels of psychological distress may experience greater role-related concerns and fewer role-related rewards. With cross-sectional data, as in this paper, it is not possible to determine the direction of causation. Under the main-effects model, the quality of experience in one role does not affect the relationship between the quality of experience in the other role and psychological distress. With respect to the effects of the employee and mother roles, the main-effects model indicates no compounding of effects from one role to the other. To illustrate, the home-based woes or joys of employed mothers would not compound the effects of their work experience on their level of psychological distress. Most researchers studying the effects of role quality on mental-health states examine only one role at a time, implicitly assuming this main-effects model. For example, studies of the relationship between work stress and health outcomes often do not assess family-role quality (see for example, Karasek, Schwartz, Theorell, Pieper, Russell, & Michela, 1982). In contrast, the interaction-effects model posits that emotional feelings generated in one arena-work or family-"spillover" into the other, increasing vulnerability or resilience to psychological distress. To demonstrate spillover effects on psychological distress, one has to show that the interaction between the experiences in the two roles accounts for a significant proportion of the variance in psychological distress, over and above that accounted for by the main effects of the experiences in the two roles. This approach
Downloaded by [New York University] at 12:25 08 January 2015
FIGURE la. Main-effects model of the relationship between job-role and parent-role quality and psychological distress.
WORK-ROLE QUALITY
PSYCHOLOGlCAL DISTRESS PARENT-ROLE QUALITY
c
Downloaded by [New York University] at 12:25 08 January 2015
Downloaded by [New York University] at 12:25 08 January 2015
Rosalind C . Barr~cttand Nancy L. Marslull
15
differs from that taken by most research on role quality across domains, which does not look at health outcomes. The typical study documents correlational relationships between affectiveexperiences in one role and affective experiences in another role. Moreover, most of these studies have been done on male samples (Evans & Bartolome, 1980, 1984) and report positive correlations between occupational stress and family dissatisfaction. In a study of men and women, Crosby (1984) reported a positive correlation between job satisfaction and home satisfaction. Such findings do not address the question of whether these correlational relationships have mentalhealth consequences. While an interaction between the quality of experiences in two roles will test the existence of spillover, we would also like to disaggregate negative spillover from positive spillover. To do this, we use two different measures, one of the positive aspects, or rewards, of a role, and the other of the negative aspects, or concerns, of a role. We then examine whether either the rewards of Role A, or the concems of Role A, interact with the overall quality of Role B. If there is an interaction between the rewards of Role A and the overall quality of Role B, we can examine a graph of this relationship to determine whether there is positive spillover from Role A to Role B, that is, whether women with rewards in Role A are at reduced risk for psychological distress compared to other women with similar experiences in Role B. Similarly, if there is an interaction between the concems of Role A and the overall quality of Role B, we can examine a graph of this relationship to determine whether there is negative spillover from Role A to Role B. We can also test for positive and negative spillover from Role B to Role A by examining whether the rewards or the concerns of Role B interact with the overaN quality of Role A. Using data from a disproportionately-stratified random sample of 403 employed mothers, ages 25 to 55, this study addresses the following questions: 1. Are employed mothers more distressed than working women without children? 2. Do women's experiences at work and as parents have main effects on their experience of psychological distress'?
WOMEN & NEALTti
16
3. Is there positive or negative spillover from one role to another that contributes to psychological distress?
METHODS
Downloaded by [New York University] at 12:25 08 January 2015
Sample The sample for this paper consists of 403 employed women, who were part of a larger study of employed women. The data come from the first wave of a three-wave longitudinal study (over two years) of a disproportionately-stratified random sample of 403 Massachusetts women employed at least half-time in one of two health-care professions-licensed practical nursing and social work. We selected these two professions because (1) they are predominantly female. (2) they produce high job strain? and (3) complete public licensure records are available. Since the Commonwealth of Massachusetts requires registration as a condition of practicing these professions, the professional registries comprise the entire population of registered social workers and LPNs in the state. Within each occupation, the sample was stratified by race (White or Black), parental status (has a child or does not have a child) and partnership status (marriedfliving with a partner or not being in either type of relationship). Because of the low percentage of Black employed women listed in the registries (e.g., only 4% of registered social workers were Black and employed), we oversampled Blacks to obtain a group large enough for analysis. In addition, because the role-status variables are confounded naturally in the population (e.g.. Black women 25-55 are less likely than White women 25-55 to be currently married), we further stratified the sample within occupation by race, partnership status, and parental status. The success of our stratification is reflected in the low correlations among occupation, race, partnership status and parental status. For example, among social workers, the correlation between race and partnership status is .11; between race and parental status, the correlation is -.11. The comparable figures for the LPNs are .04 and -.06.
Downloaded by [New York University] at 12:25 08 January 2015
Rosalind C. Borne11 and Nancy L. Marsl~all
17
Of the potential subjects whom we contacted, only 4% refused to participate. The total sample consisted of 403 women; 248 social workers (61.5%) and 155 licensed practical nurses (38.5%). Sixtyone women (15.3%) were Black, and 342 (84.7%) were White. The mean age of the respondents was 39.5 years (sd = 7.4). Of the 198 women who were partnered (49.1%). 75 did not have children, 123 did. Of the 205 women (50.9%) who were single, 99 did not have children, 106 did. Most of the mothers' (72%) youngest child is school-aged, that is, between the ages of 6 and 18 years; 13.9% had a child under age six, only 14% of the mothers do not have a child under the age of 19. Finally, although the average number of children was 2.5, the mean number of children currently living at home was 1.5 (sd = 1.2). On average, the respondents had been working in their respective fields for 11 years (the range was 2 to 35 years) and at their current jobs for 5.7 years (sd = 5.1). They worked 39.4 hours per week, on average (sd = 8.7). and 80% worked the same schedule on a regular basis. The mean individual income in 1985 was $24,400 (sd = $2,700). All resoondents lived within a 25-mile radius of Boston. Respondents were interviewed in their homes or offices by a trained interviewer. Interview questions were structured and responses were recorded by the interviewer, using pre-coded response categories; Interviews lasted about 2 hours and covered the quality of the woman's experiences in each of her major social roles, i.e.. employee, mother, partner, as well as indices of psychological distress, well-being, and physical health. Respondents were paid $10 for participating. Although the sample is representative of Massachusetts women employed in these two heavily female occupations, it is not representative of women employed in a wider range of occupations. It is not clear whether or how predispositions to such jobs affects the findings of the study. One possibility is that women who assume nurturant, caring jobs are less likely than women not in such jobs to compartmentalize their affective experiences. They might be expected to carry problems with their children with them to the workplace and to bring their emotional distresses home with them, increasing their vulnerability to the negative effects of stressors in their other role. Conversely, emotionally rewarding
I8
W O M M & HEALTH
experiences in one role might have special meaning to these women. thereby increasing their resilience to the effects of stressors in their other role. As a result, these women might be more likely than women not attracted to these fields to experience spillover effects, negative as well as positive, both from job to home and from home to job.
Downloaded by [New York University] at 12:25 08 January 2015
Measures Psychological distress. Psychological distress was assessed by the anxiety and depression subscales of the SCL-90-R, a frequency of symptoms measure (Derogatis. 1975; 1983). Respondents indicate on a 5-point scale (from 0 = not at all, to 4 = extremely) how often in the past week they were bothered by each of 14 symptoms of anxiety and 10 symptoms of depression. The decision to combine the scales into a psychological distress score was based on the high correlation ( r = .80) between the scales and on the similarity in the pattern of correlations between the anxiety and depression scales and the other variables of interest in the study. The use of combined scale scores follows the work of Barnett and Marshall (1991) and Folkman, Lazarus, Gruen, and DeLongis (1986). The SCL-90-R has high levels of both internal consistency and test-retest reliability. In this sample, coefficient alpha was .88 for depression and .89 for anxiety. These figures are similar to those reported by Derogatis (1983). Satisfactory test-retest correlations (.82 for depression and .80 for anxiety) have also been reported (Derogatis, 1983). Role rewards and role concerns. Positive and negative aspects of the roles of worker and mother were assessed by reward and concern scales constructed originally from data gathered during extensive interviews with 72 women, ages 35 to 55 (Baruch & Bamett, 1986 for a full discussion). These scales have been revised and expanded, drawing on psychometric data from a sample of 238 women, focus groups with an additional 30 women employed as licensed practical nurses and social workers, as well as the work of other researchers. For each role, respondents are asked to think about their situation as it is right now and to indicate on a 4-point scale (I =not at.all, to 4 = extremely) to what extent, if at all, each
Downloaded by [New York University] at 12:25 08 January 2015
Rosalind C . Barnc11 and Nancy L. Marsl~all
19
of the items is rewarding (or of concern).' (The number of items differs across roles; for the role of worker, there are 25-reward and 25-concern items; for the role of parent, there are 18-reward and 20-concern items.) For example, for the role of paid worker, each subject was asked how rewarding she found "the job security" and to what extent was "the job's not using your skills" a concern. For the role of mother, each woman with children was asked how rewarding she found "the love they show" and how'much of a concern was "how they spend their free time" (items comprising the reward and concern scales are shown in Appendix A). Each subject received two scores for each social role: a reward score and a concern score. Test-retest reliability coefficients, calculated on a 10% random subsample retested within 1-3 months of the initial interview, were ..88 for both job rewards and job concerns, .82 for parent rewards, and .70 for parent concerns. Internal consistency reliability coefficients (Cronbach alphas) for job rewards was 3 8 ; for job concerns, it was -89; for parent rewards, .83; and for parent concerns, .89. The job rewards and job concerns scores are correlated -.57. Overall role q~~ality. Overall role quality is conceptualized as the balance between the positive and negative aspects of a role, and is operationalized as the sum of the reward scores and the inverse of the concern scores for each role, adjusted for scale length. This score provides an index of the overall quality of experience women have in their job and parent roles. Previous research indicates that overall role quality is associated with psychological distress (Bamett & Baruch, 1985; Baruch & Barnett, 1986; Bamett, Davidson, & Marshall, 1991; Bamett, Kibria, Marshall, & Pleck, 1991; Bamett & Marshall, 1991). RESULTS
Preliminary analyses comparing the two occupational groups indicated no significant differences in the means for either the jobreward, job-concern, parent-reward, or parent-concern scales. Using a dummy variable for occupation, a series of regression models was
Downloaded by [New York University] at 12:25 08 January 2015
20
WOMEN & HEALTH
estimated to test for the main effects of occupation on psychological distress, as well as interactions between occupation and race, age, and per capita income. Because occupation was never a statistically significant predictor, nor did the main effect of role quality differ by occupation, the two occupational groups were combined for the analyses reported in this paper. Correlations between the reward and concern scores for the job and parent roles are shown in Table 1. As noted above, within each role, rewards and concerns are negatively correlated. Employed women who report more positive experiences at work, report fewer negative experiences at work; similarly, employed mothers who report morepositive experiences in their maternal role, report fewer concerns with their children. These moderate negative correlations between the rewarding and problematic aspects of these roles are expected, because "'good' jobs tend to be high on rewards and low on concems while 'bad' jobs tend to have problems in more than one area" (Hackman and Oldham, 1975, p. 166). ,In fact, the magnitude of the correlations between the job rewards and job concerns scores indicate that the job rewards score explains only 32% of the variance in the job concerns score; fully 68% of the variance in the job concerns score is unique to that measure. Similarly, parent rewards and parent concerns are correlated -.22; 95% of the variance in parent concems is unique to that measure. As expected, the correlations between job-role quality and psychological distress ( r = -.42, p < .001) and between parent-role quality and psychological distress (r = -.23, p < .001) were negative and significant. Employed women who report more positive experiences at work, report fewer symptoms of psychological distress. Employed mothers who report more positive experiences in their parenting role report lower levels of psychological distress. The correlations between the reward and concern scores in each role and psychological distress range between -.07, ns and .37, p < .001 and are in the expected direction (i.e., job and parent rewards are correlated negatively with distress; job and parent concerns are correlated positively). Role concern account scores for approximately 12.5% of the variance 'in psychological distress, whereas role reward scores account for between -50% and 10.0%.
21
Rosalind C. Darrieff and Namy L. Marslull
TABLE 1. In~ercorrelationsBctween Rewvd and Concern Scores for Job and Pnrenl Rolcs
Downloaded by [New York University] at 12:25 08 January 2015
Job Rewards Job Qncerns Parent Rewards parent ~ o ~ m r n s
Note. *p
-.39***
.17** -.07
.14*
-.17* .36***
-.31*** -.22**
= 211
< .05; **p < .Ol; ***g < .001.
Research Question 1: Are Employed Mothers More Distressed Than Employe'd Women Without Children? To examine the relationship between parental status and psychological distress, a dummy variable for parental status ( I = parent, 0 = non-parent) was entered as a predictor into a regression model that included a set of four controls: age, race, education, and per capita income. The sample for this analysis consisted of all 403 employed women. Although the overall regression model was significant, parental status was not. This finding indicates that among employed women, parental status per se is not a significant predictor of psychological distress; in other words, after controlling for age, race, education and per capita income, employed mothers
Downloaded by [New York University] at 12:25 08 January 2015
22
WOMEN & HEALTH
are at no greater risk of psychological distress than are employed women without children. Other factors. There are data to suggest that single mothers are at higher risk for stress-related mental-health problems than are partnered mothers (McLanahan & Adams, 1987). Firm conclusions about the reasons for the vulnerability of single mothers are questioned by recent findings indicating that after controlling statistically for the effects of per capita income, occupation, and education, employed single mothers are at no greater risk for psychological distress than are employed partnered mothers (Bamett & Marshall, 1991). Risk for depression among mothers has also been related to the number of preschool-age children at home (McLanahan & Adams, 1987; Pearlin, 1975). Thus, although the main focus in this paper is on the roles of paid employee and mother, in every analysis the effects of partnership status, number of children living at home, number of preschool children, and having children eighteen years of age or younger are also examined. In this instance, we first tested the main and interactive effects of partnership status in predicting psychological distress, trying to ascertain whether employed single mothers report higher levels of distress than employed partnered mothers. We examined two models: one with the main effects of partnership status (I = partnered, 0 = single) and another with the main effect of partnership status and the interaction between parental status and partnership status. Neither the main nor interaction effects was significant. Thus, among employed women, after considering the effects of age, race, education, and per capita income, psychological distress is unrelated to parental status, regardless of whether they are partnered or single. After examining separately the main effects of number of children at home, number of preschool children, and having children eighteen years of age or younger, we examined interactions between each of the predictors and partnership status by creating the appropriate interaction terms, (e.g., partnership status x number of preschool children, partnership status x number of children living at home). Once again, all effects were non-significant. Thus, among employed mothers, after considering the effects of age, race,
Rosalind C. Borne11 and Nancy L. Marshall
23
education and per capita income, psychological distress is unrelated to parental status, independent of whether the employed mothers were partnered or single, or had preschool-age children, had children eighteen years of age or younger, or had few or many children living at home.
Downloaded by [New York University] at 12:25 08 January 2015
Research Question 2: Do Women's Experiences at Work and as Parents Have Main Effects on Their Experience of Psychological Distress? Analyses of the relationship between psychological distress and parent-role quality and job-role quality were conducted on the 228 employed mothers4 in the sample who had complete demographic data. Not surprisingly, the subsample of employed mothers differed significantly from the total sample with respect to age and per capita'income (Table 2). The mothers were significantly older and, as a consequence of their larger family size, had lower per capita income. Employed mothers were also more likely than employed non-mothers to be LPNs than social workers, however, they did not differ on race. Fifty and nine-tenths percent (n = 116) of the employed mothers were married, 36.4% (n = 83) were divorced, 12.7% (n = 29) were never-married or separated. Not surprisingly, parent-role quality shows the predicted relationship with distress, even after controlling for the effects of mothers' age, race, education, per capita income, and job-role quality (Table 3). Similarly, job-role quality is significantly related to psychological distress, .even after controlling for age, race, education, per capita income and parent-role quality. The regression equation was significant, (F (6, 204) = 4.77, p < .001, RZ = .12. While psychological distress is unrelated to parental status, employed mothers with positive mother-child experiences report lower levels of psychological distress, and employed mothers with troubled mother-child relationships report higher levels. Thus, the quality of the maternal role, not the occupancy of the role, predicts distress among employed mothers.
WOMEN & HEALTH
24
Downloaded by [New York University] at 12:25 08 January 2015
TABLE 2. Descriptive Data Comparing Employed Mothers to Employed Nonmothers
42.7***
6.43
35.5
15,559***
9301
27,397
ompatima
1.56**
.50
1.69
.46
~aceb
1.11
.32
1.18
.30
Age
Per Capita Imome
6.61
U,324
a 1 = Licaszd Fractical Nurse; 2 = Social Wcnksr b 1 = Wni-;
2 = Black
Research Question 3: Is There Positive or Negative Spillover from One Role to Another That Contributes to Psychological Distress? Consistent with previous research, we found a significant positive correlation ( r = .29, p < .001)between parent-role quality and jobrole quality. Employed mothers who report higher levels of positive-role quality in one role tend to report higher levels of positive-role quality in their other role. However, any one role explains less than 10% of the variance in the other role. We then examined whether these associations between the rolequality measures translate into spillover effects. operationalized as the interaction between the measures in predicting psychological
Rosalind C. Bartiell and Nancy L. Morslwll
25
TABLE 3. ~ a i n - ~ f f i cModel: ts Relationship of Parent-Role and Job-Role Quality to Psychological Distress
we
Downloaded by [New York University] at 12:25 08 January 2015
Race
EcWation Per Capita IKIlmne
Parent-Role Qality Job-Role puality
Note.
= 211
a Unstmlardlzed regression xefficients b Startiard error
of the unstandardized regression coefficient
c Standardized regression caeffioents p e0.05; *%
