Journal of Personality and Social Psychology
1986, Vol. 51, No. 3,578-585
Copyright 1986 by the American Psychological Association, Inc.
0022-3514/86/$S0.75
Role Quality, Multiple Role Involvement, and Psychological
Well-Being in Midlife Women
Grace K. Baruch and Rosalind Barnett
Brandeis University
Women's occupancy of the social roles of paid worker, wife, and mother, and the quality of their
experience in these three roles, were examined in relation to psychological well-being. Data were
from a disproportionate random sample (N — 238) of Caucasian women ages 35 to 55. Well-being
was measured by indices of self-esteem, depression, and pleasure; pleasure was assessed by a scale
consisting of single-item measures of happiness, satisfaction, and optimism. Role quality was measured by scales developed for this study that assessed the balance between the positive and negative
attributes women perceived in their roles. Hierarchical regression analyses controlling for age, education, and income indicated that role occupancy per se was unrelated to well-being with one exception: occupying the role of paid worker significantly predicted self-esteem. In contrast, the three role
quality variables were significant predictors of the well-being indices, with one exception: quality of
experience in the role of mother did not predict pleasure. Overall, the findings suggest the importance of qualitative rather than quantitative aspects of role involvement and the need to examine
different dimensions of well-being in relation to social roles.
involvement in several roles is seen as yielding a variety of
In a review of the literature on women's social roles, Long
sources of stimulation, gratification, and social validation.
and Porter (1984) point to the existence of a kind of sex segregation in research and theory on multiple roles. For men, multiple
Recent empirical evidence and theory are converging to sug-
roles have been seen as beneficial; Gove and Tudor (1973), for
gest that for women as well as men, involvement in a multiplic-
example, attributed the better mental health of men compared
ity of roles yields a net gain of benefits over costs with respect
to women to their participation in both family and paid work
to both physical and mental health. These studies tend to sup-
roles. The topic of "dual roles," in contrast, has been seen as a
port the enhancement hypothesis, because findings for both
woman's issue; for women, the role of paid worker is assumed
women and men typically show a positive relation between the
to be added on to the normative condition of being a wife,
number of roles a person occupies and various indices of psy-
mother, and homemaker. Role overload, role conflict, guilt, anx-
chological well-being (Crosby, 1983; Epstein, 1983; Pietromo-
iety, and other hazards are expected to follow, resulting in im-
noco, Manis, & Frohardt-Lane, 1984; Thoits, 1983; Verbrugge,
paired well-being.
1982). For example, Pietromonoco et al. (1984) report that the
Two major hypotheses have been put forward concerning the
more roles a woman occupies, the more different areas of her
relation of role involvement to well-being. The "scarcity" hy-
life she reports to be sources of pleasure.
pothesis (Marks, 1977), put forth by Goode (1960) and others,
Both the enhancement and scarcity hypotheses are limited,
assumes that the social structure normally creates overly de-
however, by their focus on the number of roles occupied; both
manding role obligations, the more so the more roles one occu-
fail to examine how the nature of a particular role might con-
pies. Because human energy is limited, well-being is impaired
tribute to or impair well-being. The expansion hypothesis, for
by the overload and conflict inherent in numerous, often incom-
example, assumes a net gain of benefits over costs regardless
patible roles. In contrast to this view, the "enhancement" hy-
of which roles a person occupies. It may be, however, that the
pothesis (Marks, 1977; Sieber, 1974) emphasizes the benefits
particular roles occupied, and the quality of experience in each
rather than the costs of multiple role involvement: status, privi-
role, affect level of well-being more than the mere number of
leges, increased self-esteem, the ability to trade off undesirable
roles. The privileges and obligations, the rewards and concerns,
components of roles. Like placing one's eggs in many baskets,
and the cost/benefit balance for a woman who occupies the two
roles of wife and mother may differ from those for a woman who
occupies the two roles of wife and paid worker. Moreover, two
women may occupy similar roles yet experience the quality of
The research reported here was supported by National Science Foundation Grant BNS 77-26756. The authors extend their appreciation to
Erin Phelps and Nancy Marshall for their assistance in data analysis and
to Nathalie Dana Thompson for her assistance in manuscript preparation.
Correspondence concerning this article should be addressed to Grace
K. Baruch, who is now at the Wellesley College Center for Research on
Women, Wellesley, Massachusetts 02181.
each role differently.
Another issue in examining the relation of well-being to role
involvement is that the way well-being is conceptualized and
operationalized affects the findings. Psychological well-being is
not a unitary construct (Andrews & Withey, 1974; Bryant &
Veroff, 1982; Campbell, Converse, & Rodgers, 1976), and studies of social roles and well-being have focused on varying as-
578
ROLE QUALITY IN WOMEN
pects, such as self-esteem, happiness, satisfaction, or symptomatology, both singly and in combination. As the following brief
review indicates, a clearer understanding of the relation of role
involvement to well-being requires examining specific roles and
their quality, in relation to specific aspects of well-being.
Paid Worker
Evidence is growing that employment typically has positive
psychological consequences for women, either as a primary
source of well-being or as a buffer against stress experienced in
other roles (Barnett& Baruch, 1985; Baruch, Barnett, & Rivers,
1985; Belle, 1982; Brown & Harris, 1978; Coleman & Antonucci, 1982; Crosby, 1983; Epstein, 1983; Kessler & McRae,
1981;Merikangas, 1985; Veroff, Douvan, &Kulka, 1981). Correlational studies such as these of course raise the question of
the impact of selection: Are women selected into employment
because of their high level of well-being? If so, well-being has
"caused" them to be employed rather than vice versa. In a careful analysis of this question, Vebrugge (in press) has concluded
that selection factors account for only a small proportion of the
findings. Involvement in the role of paid worker, even if it requires that a woman must juggle many roles, appears to provide
self-esteem and a sense of efficacy, alleviating effects of stressors.
For example, a classic study of English women showed that for
those in stressful life circumstances who did not have confidantes, being employed seemed to protect against the occurrence of psychiatric symptoms; symptoms developed in 79% of
the nonemployed women but only 14% of the employed (Brown
& Harris, 1978). More recently, Merikangas (1985) reported,
based on data from a large-scale cachement area study of the
incidence of psychiatric problems, that lack of employment is
a risk factor for depression in women. Indeed, Long and Porter
(1984) suggest that for women, the benefits of multiple role involvement are in fact due primarily to the paid worker role.
These benefits, the studies just cited suggest, have a positive impact on well-being primarily with respect to self-esteem, and to
a lesser extent, psychiatric symptomology. When happiness or
satisfaction are used as indices of well-being, employed and
nonemployed women typically do not differ significantly
(Campbell et al., 1976; Serlin, 1980). These findings, however,
refer to the effects of employment per se and do not take into
account qualitative aspects of the paid worker role.
Wife
The heated debate about whether marriage is a "health hazard" for women reflects conflicting evidence about the relation
of marital status to well-being (Aneshensel, Frerichs, & Clark,
1981; Bernard, 1972; Campbell etal., 1976; Glenn, 1975). One
of the most consistent findings is that married women are happier and more satisfied than are those who are not married,
whether the latter are never-married, divorced, or widowed
(Baruch et al., 1984; Campbell et al., 1976; Depner, 1979;
Glenn, 1975; Veroff etal., 1981; Ward, 1979). A recent review
using meta-analytic techniques, however, showed that associations between subjective well-being, for example, happiness and
morale, and marital status tend to be weak (Haring-Hidore,
Stock, Okin, & Witter, 1985). In contrast, occupying the role of
579
wife typically is not associated with components of well-being
other than happiness and satisfaction. Comparing never-married women over 30 with a matched sample of married women,
Gigy (1980) found that never-married women were less happy,
but their self-esteem was no lower nor did they report more
psychiatric symptomatology. Indeed, Cleary and Mechanic
(1983) have noted that gender differences in psychiatric symptomatology are greatest among married subjects. Although being married per se may typically enhance feelings of happiness
and satisfaction, the effect is weak, and marital status apparently neither boosts women's self-esteem nor ensures against
the development of psychiatric symptomatology. When the
quality of the marriage is taken into account, however, the impact on well-being is stronger. For example, consistently negative interactions with the husband strongly predicts depression
in wives (Brown, 1984.)
Mother
In contrast to the findings with respect to the roles of paid
worker and wife, there is little evidence that being a mother per
se enhances any aspect of a woman's well-being (Barnett & Baruch, 1985; Campbell et al., 1976; Depner, 1979; Sears & Barbee, 1977; Veroff etal., 1981).
That such a seemingly profound difference in women's
lives—that between having children and being childless—
should be undetectable in its effects on well-being is counterintuitive and somewhat startling. Yet there is at least one obvious
reason all too familiar to parents: children can make one's life
miserable as well as wonderful. If one groups together all women
who happen to be mothers, those for whom it has worked out
well may be balanced by those who have had more suffering
than pleasure from their role. In Occupation Housewife, Helena
Lopata (1971) argues that being a mother is made especially
difficult in our culture by the lack of agreement on standards
for the role, combined with the unrealistic expectation that one
can mold a child to be whatever one wants. Veroffet a). (1981)
attribute the lesser well-being of women compared to men to
the difficulties and uncertainties of having primary responsibility for childrearing, and to feeling responsible for how children
turn out, despite the inability to control the outcome. According to Spreitzer, Snyder, and Larson (1979) the hypothesis that a
good experience in one role compensates for problems in other
roles is generally supported, with one exception—for women,
problems in the parent role are not compensated for by satisfaction at work. Overall, however, the relation between the quality
of a woman's experience in the role of mother and specific dimensions of her psychosocial well-being remains unclear.
As this brief review indicates, roles are not interchangeable
in their impact. In analyzing the relation of multiple role involvement to well-being, we need to examine not only the quantity but the quality of a woman's experience in specific roles.
The data reported here examine the relative contribution of
women's occupancy of social roles and the quality of their experience within roles to three dimensions of psychological wellbeing.
With respect to well-being, Bryant and Veroff( 1982) have recently presented a three-dimensional model based on factor
analyses of data on 18 indices of well-being included in two
580
GRACE K. BARUCH AND ROSALIND BARNETT
Table 1
Role Pattern Groups
Work status
Employed
n
Not employed
n
Never
married
Married
without
children
Married
with
children
Divorced
with
children
49
40
46
46
15"
42
Note. N = 238.
" It was not possible to locate the desired number of married, childless,
non-employed women. Economic pressures and ideological changes
have apparently made this role pattern very rare.
large-scale national surveys. In their model (negative poles are
labeled), they distinguish personal inadequacy (poor self-concept), strain (symptomatology), and unhappiness (negative
affect). The study reported here also included multiple measures of well-being. Three indices analogous to the Bryant and
Veroff dimensions were selected for analysis, self-esteem
(Ro-
senberg, 1965), depression (Derogatis, Lipman, Rickels, Uhlenhuth, & Covi, 1974), and pleasure (a scale constructed of three
standard single-item measures of happiness, satisfaction, and
optimism).
With respect to role quality, positive (rewarding) and negative
(distressing) role attributes were measured so as to assess the
difference or balance between the two. The balance between rewarding and distressing aspects of one's roles seemed a key to
of women who occupy one (n = 64), two (n = 126), or three (n = 46) of
the major roles of wife, mother, paid worker.
The first stage of the study consisted of intensive, semistructured interviews lasting 3 to 6 hr with a "snowball" or convenience sample of
12 women from each of the six groups shown in Table 1 (JV = 72). A
major goal of these interviews was to identify the rewarding and distressing aspects of roles so as to assess role quality. Based upon content analyses of transcriptions of these taped interviews, a survey instrument was
designed for the second stage of the study, in which approximately 45
women in each of the six role pattern groups were selected through disproportionate random sampling, using the voting lists of a community
in the greater Boston area. Details of sample recruitment are provided
in Baruch et al. (1985). Briefly, census data were used to assess the characteristics of communities in the Boston Standard Metropolitan Sampling Area (SMSA) so as to locate the community most likely to have
adequate numbers of employed women and a wide range of occupational prestige levels. Based on that community's voting list, every
woman aged 35 to 55 was assigned a random number; subjects were
contacted in order of number. Telephone screening was used to assess
whether a woman met the sampling criteria. (Although 6,000 women in
the community were screened, it was not possible to locate adequate
numbers of married childless women who were nonemployed. The
changing economic and ideological climate has apparently made this
pattern very rare.) Women were interviewed individually in their homes
in 1978-1979; the response rate was 76%.
The mean age of the sample was 43.6 years. The mean level of education was 14 years, 2 years beyond high school; the mean total family
i ncome for the whole sample was $21,600.2 Subgroups varied greatly in
income; however, the mean for never-married women was $15,200; for
employed married women it was $30,700. In the sample as a whole, 180
of 238 women were employed, 142 were married, and 136 were mothers. Of the mothers, 30 had children under 8, and 89 had children 8 to
18 living at home.
the quality of that role in relation to mental health; several researchers have argued, for example, that the balance between
positive and negative affect is the best predictor of one's psychological state (Bradburn, 1969;Lowenthal&Chiriboga, 1973).'
Method
Subjects
The sample, described in Table 1, consisted of women who occupied
one of four family role patterns: never-married, married without children, married with children, and divorced with children. Half of the
married women and all of the never-married and divorced women were
employed. Employed women within each family status group were
drawn equally from those in high-, medium-, and low-prestige occupations (Siegel, 1971). Nonemployed women were stratified similarly by
the prestige level of their husbands' occupation. Women who had been
working at least 17 hr a week for at least 3 months prior to being interviewed were classified as employed. Divorced women had been divorced
or legally separated for at least a year prior to being interviewed. All
subjects were Caucasian.
The sample was selected to include (a) groups of theoretical relevance
to the major concerns of the study—for example, groups vary in the
number and nature of roles—and (b) relatively rare groups whose life
patterns are of increasing social importance, such as married childless
women and women in high-prestige occupations. The study was designed to sample specific groups rather than to be a probability sample
of American women. Employed women in high-prestige occupations,
for example, are overrepresented; one purpose of the design was to
avoid confounding the effects of being employed with the effects of being
employed in a low-prestige occupation. The design yields six subgroups
Measures
Psychological well-being. Three indices of well-being included in the
survey were selected for final analyses, based upon the conceptual and
empirical considerations discussed earlier. These were the Rosenberg
(1965) Self-Esteem Scale; the depression subscale of the Hopkins Symptom Checklist (Derogatis et al., 1974), and a pleasure scale consisting
of three standard single-item measures of happiness, satisfaction, and
optimism. The alpha coefficient for the pleasure scale was .76.
Role quality. As noted earlier, 72 women (12 per role pattern group)
were interviewed in the pilot stage of the study to identify the rewarding
and distressing aspects of each role they occupied. On the basis of re-
1
Our approach to assessing the quality of experience in a role also
builds upon recent research suggesting that levels of psychological wellbeing and stress are affected more by "chronic" difficulties and gratifications inherent in social roles than by major, discrete life events (Pearlin, 1982). Daily concerns have been related to psychiatric disfunction,
including anxiety and depression (Kanner, Coyne, Schaefer, & Lazarus,
1981; Lewinsohn & Talkington, 1979). Festively toned experiences,
namely, rewards, are viewed as important both in preventing and mitigating the effects of stress (DeLongis, Coyne, Dakof, Folkman, & Lazarus, 1982; Kanner etal., 1981).
2
Family income was classified as follows: I = $1.00-$3,999; 2 =
$4,000-$5,999; 3 = $6,000-$9,999; 4 = $10,000-$15,999; 5 =
$16,000-$20,999; 6 - $21,000-$28,999; 7 = $29,000-$39,999; 8 =
$40,000-$64,999; 9 = $65,000 and above. Education was categorized
as follows: 1 = some high school or less; 2 = high school diploma; 3 =
some college or trade school (without degree); 4 = trade school with
degree; 5 = college degree; 6 = M.A.; 7 = higher than M. A.
ROLE QUALITY IN WOMEN
581
Table 2
Well-Being Indices: Means and Standard Deviations for the Six Role Pattern Groups
Self-esteem
Depression
Pleasure
Roles occupied: Group"
M
SD
M
SD
M
SD
Wife(M,NC,NE) a
Paid worker (NM)
Wife, mother (M,C,NE)
Paid worker, wife (M, E, NC)
Paid worker, mother (D)
Paid worker, wife, mother (M, C, E)
3.35
3.37
3.30
3.57
3.56
3.59
.49
.57
.59
.44
.36
.41
.48
.48
.35
.38
.38
.33
.27
.45
.34
.49
.36
.29
-.69
-.94
-.45
2.48
2.58
2.62
2.61
1.94
2.46
.28
-.05
.72
Note. The means and standard deviations for the total sample were as follows: for self-esteem, M = 3.47, SD = 0.49; for depression, M = 1.39, SD =
0.39; for pleasure, M = .03, SD = 2.46.
' Initials refer to the sampling groups in Table I as follows: C = children; D = divorced; E = employed; NC = no children; NE = not employed;
NM = never married.
sponse frequency, equal number of rewards and concerns were used to
construct scales for subjects in the survey stage of the study. Subjects
used a 4-point scale to indicate to what extent, if at all, each of the items
was rewarding or distressing. For example, for the role of paid worker,
each employed subject was asked how rewarding she found the pay she
received and how much of a concern a lack of challenge was. Each subject received three scores per role: a mean reward score (her mean response to the reward items), a mean concern score, and a balance score,
for example, the difference between the mean reward score and the
mean concern score. The balance scores constituted our index of the
quality of experience in each role.
reliability, alpha coefficients were calculated for the six scales
and ranged from .71 to .94.
Women who occupied the role of paid worker reported as the
most rewarding aspects of their jobs being able to work on one's
own, the sense ojaccomplishment/competence, and having a job
that fits one's interests and skills. The most distressing aspects
were having too much to do, having to juggle conflicting tasks,
and lacking career growth opportunity. For the role of wife, the
most rewarding aspects were having a husband who backs you
up in what you want to do, the companionship, and having a
husband who sees you as special. The most distressing aspects
were husband's physical health problems, conflicts over chil-
Results
dren, and husband's job problems. For the role of mother,
women found the most rewarding aspects to be the love children
Intercorrelations Between the Well-Being Variables
show, liking the kind of people they are, and pleasure in their
accomplishments. The most distressing aspects were worry
Intercorrelations between the well-being variables were
highly significant. Self-esteem was correlated with depression
at -.63
(p < .001) and with pleasure at .51 (p < .001). Depres-
sion was correlated with pleasure at —.61 (p < .001). Thus the
three indices measure related dimensions of well-being; each
Table 3
Reward Items for Paid Worker Role: Means, Standard
Deviations, and Correlations With Well-Being
explains from 26% to 40% of the variance of the others.
Role Occupancy and Weil-Being: Correlational Analyses
The number of roles a woman occupied was significantly correlated with the well-being indices; for self-esteem, the correlation was .16 (p < .01); for depression it was -.14 (p < .05); and
for pleasure it was .23 (p < .001). Thus there was no evidence
that involvement in multiple roles per se was detrimental to
well-being.
Table 2 shows the mean scores on the three well-being indices
for the sample as a whole and for the six role-pattern groups.
Role Quality and Well-Being
Rewards and concerns. The reward and concern items (with
Item
M
SD
1. Hours fit your needs
2. Job security
3. Appreciation and recognition
4. People you work with
5. Helping others/being needed
6. Liking your boss
7. Sense of accomplishment/competence
8. Variety of tasks
9. Opportunity for learning
10. Physical conditions
1 1 . Getting out of the house
12. Being able to work on your own
1 3. Helping others develop
14. Job fits interests and skills
15. Good income
16. Good support facilities
17. Opportunity for advancement
18. Challenging, stimulating work
19. Getting to make decisions
2.98
2.86
3.06
3.15
3.17
2.82
3.47
3.13
3.02
2.50
2.98
3.57
3.04
3.23
2.66
2.79
2.08
3.03
3.13
0.96
1.06
0.90
0.90
0.96
1.05
0.83
0.93
1.06
1.05
1.10
0.75
1.13
0.92
1.02
0.99
1.07
1.02
0.96
means and standard deviations) for the roles of paid worker,
wife, and mother are presented in Tables 3 through 8. To assess
Note.N= 180.
582
GRACE K. BARUCH AND ROSALIND BARNETT
Table 4
Table 6
Concern Items for Paid Worker Role: Means. Standard
Concern Items for Wife Role: Means, Standard Deviations,
Deviations, and Correlations With Well-Being
and Correlations With Well-Being
Item
M
SD
Item
M
SD
1. Having loo much to do
2. Job insecurity
3. Job conflicts with other responsibilities
4. Not liking boss
5. Having to juggle conflicting tasks
6. Not getting advancement you want/deserve
7. Job's not fitting skills/interests
8. Job is too regimented
9. Bad physical conditions
10. Lack of recognition/appreciation
1 1 . Job's dullness/monotony
12. Dissatisfaction with income
13. Problems re being a woman
14. Having to do things not part of job
15. Lack of opportunity for career growth
16. Unnecessary busy work
17. Lack of challenge
18. People you work with
19. Job too draining
2.16
1.59
1.73
1.38
2.04
1.82
1.45
1.24
1.53
1.50
1.40
1.95
1.40
1.56
1.95
1.67
1.57
1.53
1.78
1.04
0.88
0.85
0.77
0.99
1.02
0.84
0.64
0.81
0.77
0.78
1.01
0.78
0.82
1.10
0.88
0.92
0.79
0.83
1 . Husband being unavailable
2. Poor communication
3. Husband's physical health
4. Not getting enough appreciation
5. Conflicts about children"
6. Husband's job/career problems
7. Problems in sexual relationship
8. Lack of companionship
9. Husband's job instability
10. Problems re demands of husband's job
11. Husband has emotional problems
12. Not getting along
13. Conflict over housework
14. Not getting enough emotional support
15. Conflicts re children
1.57
1.63
2.08
1.51
1.89
1.70
1.51
1.26
1.43
1.51
1.46
1.44
1.23
1.41
1.39
0.89
0.83
1.18
0.66
0.99
0.98
0.85
0.57
0.82
0.85
0.77
0.66
0.53
0.68
0.68
Note.N= 142.
a
Correlations not available.
Note.N= 180.
striking was that between the marriage balance score and pleasure (r = .70).
about their physical well-being, worry about the teenage years,
and the financial strain.
Regression Analyses
Balance scores and well-being. In Table 9, we present the
In the next analyses, we examined whether the quality of a
mean balance (role quality) scores for the roles of paid worker,
woman's experience in her roles, assessed by role balance
wife, and mother for the sample as a whole and for the six role
scores, added significantly to the predictions of her well-being
pattern groups. (Reward scores were consistently higher than
beyond the effects of role occupancy per se. Because relations
concern scores for all the roles assessed, resulting in positive
between well-being and the role variables could be an artifact
mean balance scores.)
of a woman's income or her educational level, the hierarchical
The relation between the balance scores and the well-being
multiple regression equations controlled for the effects of these
indices are shown in the zero-order correlations presented in
variables. (In our sample, employment status was significantly
Table 10. All the correlations were highly significant; the most
correlated with education, r= .25, p< .001, and marital status
was significantly correlated with income, r = .46, p < .001.) Age
was also entered as a control in these analyses. Role occupancy
Table 5
Reward Items for Wife Role: Means, Standard Deviations,
Table 7
and Correlations With Well-Being
Item
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
1 2.
1 3.
14.
1 5.
Companionship
Having someone to take care of you
Husband easy to get along with
Physical affection
Husband being a good father"
Able to go to husband with problems
The sexual relationship
Husband backing you up
Enjoyment of doing things for husband
Husband sees you as special
Husband is a good provider
Husband's personality fits yours
Husband's willingness to share housework
Good communication
Husband's willingness to have children
Note.N= 142.
" Correlations not available.
Reward Items for Mother Role: Means, Standard Deviations,
M
3.48
2.91
3.20
3.27
3.43
3.20
3.05
3.45
3.27
3.44
3.22
3.00
2.70
3.15
3.27
SD
0.77
0.99
0.95
0.89
0.89
0.77
0.99
1.02
0.81
0.77
0.77
0.92
0.93
1.22
0.92
and Correlations With Weil-Being and Pleasure
Item
1 . Being needed
2. Pleasure from their accomplishments
3. Helping them develop
4. The love they show
5. Feeling proud of how they are turning out
6. Liking the kind of people they are
7. Being able to go to them with problems
8. Enjoying doing things with them
9. The help they give you
10. The meaning they give your life
1 1 . Being the best caretaker for them
12. The way they get along together
13. Seeing them mature and change
1 4. The way they change you for the better
Nate.N= 136.
M
2.98
3.63
3.61
3.72
3.61
3.67
2.58
3.43
2.59
3.38
2.90
2.94
3.60
3.10
SD
0.81
0.58
0.66
0.53
0.65
0.59
1.03
0.73
0.99
0.76
1.08
0.88
0.58
0.99
ROLE QUALITY IN WOMEN
Table 8
Concern Items for Mother Role: Means, Standard Deviations,
and Correlations With Well-Being
Item
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
The financial strain
Feeling trapped/bored
Worry: physical well-being
Not getting along with each other
Heavy demands/responsibilities
Worry: teenage years
Not sure if you're doing the right thing
Their not showing appreciation/love
Problem with their education/school
Disappointment in what they are like
Not having enough control over them
Needing you less as they get older
Too many arguments/conflicts with them
Interference in relationship with husband
M
SD
2.20
1.44
2.37
1.96
2.17
2.36
2.05
1.54
1.99
1.32
1.52
1.31
1.55
1.34
0.97
0.90
1.07
1.03
.95
1.05
0.92
0.76
0.97
0.58
0.74
0.59
0.78
0.57
Note.N= 136.
variables were entered in the second step of the equations; the
balance scores, converted to z scores, were entered in the third
and last step.'
As is shown in Table 11, the total models were highly significant for all three well-being measures. For self-esteem, F(9,
193) = 6.18, p < .001; for depression, F(9, 193) = 6.54, p <
.001; and for pleasure, F(9,176) = 16.99, p < .001.
The regression analyses showed that family income was a significant predictor of all three well-being indices. The only role
occupancy variable that was a significant predictor of well-being when age, income, and education were controlled was paid
worker, which predicted self-esteem (B = .20, p < .01).
For all three well-being measures, the role quality variables
entered in the final step yielded a highly significant increment
in the percent of variance explained. For self-esteem, the most
powerful predictors after family income were occupancy of the
paid worker role (B = .20, p < .01) and the balance score for the
paid worker role (B = .20, p < .01). The two other balance scores
were also significant predictors of self-esteem. For depression,
Table 9
Mean Balance Scores of Role Pattern Groups
Role
Group
Paid worker
Never-married
Married, no children, not
employed
Married, children, not
employed
Married, employed, no
children
Divorced
Married, children, employed
1.16
Wife
Mother
1.69
1.58
1.53
1.29
1.46
1.57
1.61
1.80
1.33
1.52
Note. Mean balance scores for the total sample were as follows: paid
worker role = 1.35, SD = 0.81; wife role = 1.67, SD = 0.92; and mother
role= 1.47,S» = 0.71.
583
Table 10
Zero-Order Correlations of Mean Balance Scores
and Well-Being Indices
Well-being index
Role balance
Work
Marriage
Children
Self-esteem
r
n
.36*
.35*
.34*
178
140
132
Depression
r
n
-.35*
-.45*
-.42*
178
141
133
Pleasure
r
n
.49*
.70*
.34*
164
132
123
*p<.001.
the most powerful (negative) predictors after family income
were the balance scores for paid worker (B = -.23) and wife
(B = -.23), both significant at the p < .001 level. For pleasure,
in contrast, the most powerful predictor was the quality of experience in the role of wife (B = .47, p < .001). Family income
(B= .34, p < .001) and the balance scores for the paid worker
role (B = .31, p < .001) were also significant predictors of pleasure. In contrast to findings for the other two well-being variables, the balance score for the role of mother was not a significant predictor of pleasure.
Discussion
The overall findings of this study support the view that neither the scarcity hypothesis nor the enhancement hypothesis is
an adequate explanation of women's experiences in their social
roles in relation to well-being, in part because both hypotheses
focus on the quantity of roles.
The findings reported here suggest two important points
about role involvement and well-being. First, it is the qualitative
rather than the quantitative aspects of a woman's experiences
in her social roles that are the best key to understanding her
psychological well-being, or lack thereof. In the regression analyses, only one role occupancy variable was related to a wellbeing index; being a paid worker was associated with higher selfesteem, consistent with the literature on women's employment
and well-being just cited. Occupancy of the role of wife or
mother is not by itself predictive of well-being. In contrast, the
only role quality variable not related to all three well-being
measures was the balance score for the role of mother, which
did not predict pleasure. These findings on well-being are consistent with others focusing on stress, as assessed by indices of
role overload, role conflict, and anxiety. Analyses carried out by
the authors using the same data set indicated that role quality
variables were much stronger predictors of stress outcomes than
were role occupancy variables (Barnett & Baruch, 1985).
In their roles as workers, spouses, and parents, women (and
men) experience both suffering and gratification. Our findings
suggest that to the degree that a particular role yields a net gain
of benefits over costs, involvement in that role will have a positive impact on well-being, even if such involvement also in-
3
For each role, balance scores were set at zero for subjects not occupying that role.
584
GRACE K. BARUCH AND ROSALIND BARNETT
Table 11
strongly on their lives, the quality of work may be an even
Hierarchical Multiple Regression A nalyses:
stronger predictor of well-being. Their greater difficulties in es-
Well-Beine Indices
tablishing a stable marriage because of the economic and employment problems of Black men may make marriage and mar-
Predictor
Self-esteem
Depression
B"
B"
ital quality less or more salient to their well-being. Further, the
Pleasure
role of mother may be more central to the self-esteem of Black
women than Caucasian women (Malson, 1982). In short, re-
Control variables: Step 1
Education
Age
Family income
R2 for Stepl
Role occupancy variables:
Step 2
Paid worker
Wife
Mother
R2 for Step 2
Quality of experience
variables: Step 3"
Paid worker
Wife
Mother
R2 for Step 3
lations between well-being and involvement in social roles may
.04
-.04
.25***
.07
.07
-.08
-.24***
.06
.11
.02
.34***
.11
vary not only for different roles and different dimensions of
well-being, but for different groups of women.
References
.20***
-.05
.06
.11
.20**
.16*
.18**
.22**
-.03
-.07
-.10
.07
-.23***
-.23***
-.19**
.23**
.05
-.01
.13
.13
.31***
.47***
.09
.46***
a
Standardized regression coefficient for variable entered into regression
at that step. "Delta R1 calculations indicated that this step produced a
significant increment in variance explained.
* p < .05. **/> < .01. ***p< .001.
creases the number of roles a woman occupies. Although
merely occupying a role in no way guarantees that the role will
be rewarding, not occupying a role clearly preludes deriving any
benefits from it.
The second point is that differential patterns exist with respect to the relation of role quality to specific dimensions of
well-being. For example, a satisfying marriage was strongly tied
to pleasure (happiness, satisfaction, and optimism) but was not
as strongly associated with self-esteem or depression.
Putting these two points together, if well-being is conceptualized as having more than one dimension, and if the sources of
well-being differ for each dimension, involvement in multiple
roles may be a prerequisite for well-being.
A limitation of the findings just noted is that in a cross-sectional study, the direction of relations found cannot be known.
It is possible, for example, that women who are high in selfesteem shape their roles so as to make them satisfying. Furthermore, pre-existing emotional problems, that is, impaired wellbeing, may cause a woman to perceive her roles as low in quality.
Additional limitations of the study that restrict the generalizability of the findings include the restricted age range of the
sample and the focus on Caucasian women. With respect to age,
for example, women younger than 35 who are mothers are more
likely to have small children, and the quality of the mother role
may be a more powerful influence on well-being than for older
women. The issues addressed in this study need to be investigated in different age groups.
Perhaps most important, major racial and/or ethnic differences may exist among women. For example, because Black
women have more consistently assumed a lifelong need to be
employed, the paid worker role may relate differently to their
well-being; to the extent that the work role impacts more
Andrews, F. M., & Withey, S. B. (1974). Developing measures of perceived life quality: Results from several national surveys. Social Indicators Research, I, 1-26.
Aneshensel, C. A., Frerichs, R. R., & Clark, V. A. (1981). Family roles
and sex differences in depression. Journal of Health and Social Behavior. 22, 379-393.
Barnett, G. C., & Baruch, O. K. (1985). Women's involvement in multiple roles, role strain, and psychological distress. Journal of Personality
and Social Psychology, 49, 135-145.
Baruch, G., Barnett, R., & Rivers, C. (1985). Lifeprints: New patterns
of love and work for today's women. New York: Signet Books.
Belle, D. (Ed.). (1982). Lives in stress: Women and depression. Beverly
Hills, CA: Sage.
Bernard, J. (1972). The future of marriage. New York: World-Times.
Bradburn, N. (1969). The structure of well-being. Chicago: Aldine.
Brown, G. W, & Harris, T. (1978). Social origins of depression: A study
of psychiatric disorder in women. New York: Free Press.
Brown, G. (1984). Stress, social supports and depression- Paper presented at Harvard University School of Public Health, Boston, MA.
Bryant, F. B., & Veroff, J. (1982). The structure of well-being: A sociohistorical analysis. Journal of Personality and Social Psychology, 43,
653-673.
Campbell, A., Converse, P. E., & Rodgers, W. L. (1976). The situation
of women. In A. Campbell, P. E. Converse, & W. L. Rodgers (Eds.),
The finality of American life (pp. 395-442). New York: Russell Sage
Foundation.
Cleary, P. D., & Mechanic, D. (1983). Sex differences in psychological
distress among married people. Journal of Health and Social Behavior, 24, 111-121.
Coleman, L. M., & Antonucci, T. C. (1982, Winter). Women's wellbeing at midlife. ISR Newsletter. (Available from Institute for Social
Research, University of Michigan, Ann Arbor)
Crosby, F. (1983, November). Work and domestic life. Paper presented
at the conference "Effective Management of Work and Personal Life,"
McGill University, Montreal, Canada.
DeLongis, A., Coyne, J. C., Dakof, G., Folkman, S., & Lazarus, R. S.
(1982). Relationship of daily hassles, uplifts, and major life events to
health status. Health Psychology, 1, 119-136.
Depner, C. (1979, August). The parental role and psychological wellbeing. Paper presented at the meeting of the American Psychological
Association, New York.
Derogatis, L. R., Lipman, R. S., Rickels, K., Uhlenhuth, E. H., & Covi,
L. (1974). The Hopkins symptom checklist. In P. Pichot (Ed.), Psychological measurements in psychopharmacology. Paris: Karger.
Epstein, C. (1983, April). The new total woman. Working Woman, pp.
100-103.
Gigy, L. L. (1980). Self concept of single women. Psychology of Women
Quarterly, 5,321-340.
Glenn, N. D. (1975). The contribution of marriage to the psychological
well-being of males and females. Journal of Marriage and the Family,
37, 594-601.
ROLE QUALITY IN WOMEN
Goode, W. J. (1960). A theory of strain. American Sociological Review,
25, 483-496.
Gove, W. R., & Tudor, J. (1973). Adult sex roles and mental illness.
American Journal of Sociology, 78, 812-835.
Haring-Hidore, M., Stock, W. A., Okin, M. A., & Witter, R. A. (1985).
Marital status and well-being: A subjective synthesis. Journal of Marriage and the Family, 47, 947-954.
Kanner, A. D., Coyne, J. C., Schaefer, C., & Lazarus, R. S. (1981). Comparisons of two models of stress measurement: Daily hassles and uplifts versus major life events. Journal of Behavioral Medicine, 4, 139.
Kessler, R. C., & McRae, J. A. (1981). Trends in the relationship between sex and psychological distresses: 1957-1976. American Sociological Review, 46, 443-452.
Lewinsohn, P. M., & Talkington, J. (1979). Studies of the measurement
of unpleasant events and relationship with depression. Applied Psychological Measurement, 3, 83-102.
Long, J., & Porter, K. L. (1984). Multiple roles of midlife women: A
case for new directions in theory, research, and policy. In G. Baruch
& J. Brooks-Gunn (Eds.), Women in midlife. New York: Plenum.
Lopata, H. (1971). Occupation housewife. New York: Oxford University
Press.
Lowenthal, M. E, & Chiriboga, D. (1973). Social stress and adaptation:
Toward a life-course perspective. In C. Eisdorfer & M. P. Lawton
(Eds.), The psychology of adult development and aging. Washington,
DC: American Psychological Association.
Malson, M. (1982). Black women's sex role integration and behavior
(Working Paper #87). Wellesley, MA: Wellesley College, Center for
Research on Women.
Marks, S. R. (1977). Multiple roles and role strain: Some notes on human energy, time and commitment. American Sociological Review,
41,921-936.
Merikangas, K. (1985, May). Sex differences in depression. Paper presented at Murray Center (Radcliffe College) Conference: Mental
Health in Social Context, Cambridge, MA.
Pearlin, L. (1982, October). Social roles and personal stress. Paper pre-
585
sented at the National Conference on Social Stress Research, Durham, NH.
Pietromonaco, R., Manis, J., & Frohardt-Lane, K. (1984, August). Psychological consequences of multiple social roles. Paper presented at
the meeting of the American Psychological Association, Toronto,
Canada.
Rosenberg, M. (1965). Society and the adolescent self-image. Princeton,
NJ: Princeton University Press.
Sears, P. S., & Barbee, A. H. (1977). Careers and life satisfaction among
Terman's gifted women. In J. Stanley, W. George, & C. Solano (Eds.),
The gifted and creative Baltimore, MD: Johns Hopkins University
Press.
Serlin, E. (1980). Emptying the nest: Women in the launching stage. In
D. McGuigan (Ed.), Women's lives: New theory, research, and policy.
Ann Arbor: University of Michigan Press.
Sieber, S. D. (1974). Toward a theory of role accumulation. American
Sociological Review. 39, 567-578.
Siegel, P. M. (1971). Prestige in the American occupational structure.
Unpublished doctoral dissertation, University of Chicago.
Spreitzer, E., Snyder, E. E., & Larson, D. L. (1979). Multiple roles and
psychological well-being. Sociological Focus, 12, 141-148.
Thoits, P. A. (198 3). Multiple identities and psychological well-being: A
reformulation and test of the social isolation hypothesis. American
Sociological Review, 48, 174-187.
Verbrugge, L. M. (1982). Women's social roles and health. In P. Berman
& E. Ramey (Eds.), Women: A developmental perspective (NIH Publication No. 82-2298). Washington, DC: U.S. Government Printing
Office.
Verbrugge, L. M. (in press). Role burdens and physical health of men
and women. In F. Crosby (Ed.), Modern women: Managing the dual
role. New Haven, CT: Yale University Press.
Veroff, J., Douvan, E., &Kulka, R. (1981). The inner American: A selfportrait from 1957101976. New York: Basic Books.
Ward, R. (1979). The never-married in later life. Journal of Gerontology,
34, 861-869.
Received March 5, 1985
Revision received February 6, 1986 •