University of Southern California Dissertations and Theses

The relationship between work stress and family interaction in dual-earner couples

(USC Thesis Other)

The relationship between work stress and family interaction in dual-earner couples

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI
films the text directly from the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter face, while others may be
from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleedthrough, substandard margins,
and improper alignment can adversely afreet reproduction.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand comer and
continuing from left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in reduced
form at the back of the book.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6” x 9” black and white
photographic prints are available for any photographs or illustrations
appearing in this copy for an additional charge. Contact UMI directly to
order.
UMI
A Bell & Howell Information Company
300 North Zed) Road, Ann Arbor MI 48106-1346 USA
313/761-4700 800/521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
THE RELATIONSHIP BETWEEN
WORK STRESS AND FAMILY INTERACTION IN DUAL-EARNER COUPLES
by
Kevin Garth Flusk Thomas
A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
Master of Arts
(Psychology)
August 1996
Copyright 1996 Kevin G.F. Thomas
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 1381610
UMI Microform 1381610
Copyright 1996, by UMI Company. All rights reserved.
This microform edition is protected against unauthorized
copying under Title 17, United States Code.
UMI
300 North Zeeb Road
Ann Arbor, MI 48103
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY O F SOUTHERN CALIFORNIA
TH E GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES. CALIFORNIA § 0 0 0 7
This thesis, written by
Kevin. Garth Flusk Thomas
under the direction of h l ^ . Thesis Committee,
and approved by all its members, has been pre
sented to and accepted by the Dean of The
Graduate School, in partial fulfillment of the
requirements for the degree of
Master of Arts
D m s
n„tt A ugust 9 , 1996
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Contents
1. List of Tables/iv
2. Abstract / vi
3. Introduction /1
Spillover Effects /14
Hypothesis One /14
Hypothesis Two /15
Crossover Effects /15
Hypothesis Three /15
Hypothesis Four / 15
4. M ethod/16
Participants /16
Procedures /18
Measures /19
Adult Home Data Questionnaire /19
Work Stress / 20
Marital Interaction / 21
Intimacy/21
Conflict/21
Withdrawal / 22
Parent-Child Interaction / 23
Affection / 23
Conflict / 23
Involvement / 24
5. Results / 25
Data Analysis / 25
Descriptive Data / 30
Preliminary Data / 33
Work-to-Home Stress Contagion: The Marital Relationship / 38
Work-to-Home Stress Contagion: The Parent-Child Relationship / 40
Home-to-Work Stress Contagion: The Marital Relationship / 42
Home-to-Work Stress Contagion: The Parent-Child Relationship / 43
6. Discussion / 47
7. Notes / 50
8. References/51
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9. Appendix A: Adult Home Data Questionnaire / 56
10. Appendix B: Common Errors Sheet / 65
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
List of Tables
1. Means and Standard Deviations for Selected Couple, Spouse, and Target Child
Variables /17
2. Descriptive Statistics for Participants Daily Self-Reports of Work Stress / 31
3. Descriptive Statistics for Participants Daily Self-Reports of the Marital Relationship
/ 32
4. Descriptive Statistics for Participants Daily Self-Reports of the Parent-Child
Relationship / 33
5. Day-to-day Work Stress Contagion: Summary of Regression Analysis for Work
Stress Variables on One Day Predicting Work Stress Outcomes the Following Day
/ 34
6. Day-to-day Marital Relationship: Summary of Regression Analysis for Marital
Relationship Variables on One Day Predicting Husbands’ Experience of Marital
Interactions the Following Day / 35
7. Day-to-day Marital Relationship: Summary of Regression Analysis for Marital
Relationship Variables on One Day Predicting Wives’ Experience of Marital
Interactions the Following Day / 36
8. Day-to-day Parent-child Relationship: Summary of Regression Analysis for
Husband-child Relationship Variables on One Day Predicting Husbands’ Parent-
child Relationship the Same Day / 37
9. Day-to-day Parent-child Relationship: Summary of Regression Analysis for Wife-
child Relationship Variables on One Day Predicting Wives’ Parent-child
Relationship the Same Day / 37
10. Work-to-home Stress Contagion: Summary of Regression Analysis for Work Stress
Variables on One Day Predicting Husbands’ Marital Interactions the Same Day / 38
11. Work-to-home Stress Contagion: Summary of Regression Analysis for Work Stress
Variables on One Day Predicting Wives’ Marital Interactions the Same Day / 39
12. Work-to-home Stress Contagion: Summary of Regression Analysis for Work Stress
Variables on One Day Predicting Husbands’ Parent-child Relationship the Same Day
/ 41
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
V
13. Work-to-home Stress Contagion: Summary of Regression Analysis for Work Stress
Variables on One Day Predicting Husbands’ Parent-child Relationship the Same Day
/ 41
14. Home-to-work Stress Contagion: Summary of Regression Analysis for Variables
Within the Marital Relationship on One Day Predicting Husbands’ Work Stress
Outcomes the Following Day / 43
15. Home-to-work Stress Contagion: Summary of Regression Analysis for Variables
Within the Marital Relationship on One Day Predicting Wives’ Work Stress
Outcomes the Following Day / 44
16. Home-to-work Stress Contagion: Summary of Regression Analysis for Variables
Within the Parent-child Relationship on One Day Predicting Husbands’ Work Stress
Outcomes the Following Day / 45
17. Home-to-work Stress Contagion: Summary of Regression Analysis for Variables
Within the Parent-child Relationship on One Day Predicting Wives’ Work Stress
Outcomes the Following Day / 45
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vi
Abstract
A substantial body of research postulates that two processes, stress spillover and stress
crossover, describe the link between daily work stress and marital interactions. This study
attempted to replicate empirical studies and verify theoretical accounts of the work-family
interface. Husbands and wives from a sample of 44 dual-eamer families completed a daily
self-report questionnaire over 14 consecutive days. Questionnaire items were used to
calculate indices of work stress, marital conflict, withdrawal, and intimacy, and parent-child
affection, conflict and involvement. Data were analyzed using a pooled time series
regression. Results replicated earlier findings on the spillover process, but revealed no
crossover effects. Future longitudinal studies might explore moderating effects of the parent-
child relationship on the association between work stress and the marital relationship.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Introduction
Daily stresses at work appear to influence the employee’s emotional state and mood
at the end of the work day (Caspi, Bolger, & Eckenrode, 1987; Eckenrode, 1984; Hedges,
Jandorf, & Stone, 1985).1 A substantial body of literature supports this assertion (for a
review, see Kanter, 1977). Recent advances in methodology have enabled researchers to
document some processes that may influence the transfer of work-induced psychological
states to the home. These states may then influence the employee’s behavior, family
interactions, and the quality of spousal and parent-child relationships (Crouter, Perry-
Jenkins, Huston, & Crawford, 1989; Nesselroade, 1988).
The present study explores these processes by examining the daily variability in, and
the bidirectional links between, work stress and family interactions in dual-earner couples
and their children. An investigation bearing these characteristics is both relevant and timely
for three primary reasons: (a) Dual-earner couples now outnumber single-earner couples in
this country (Chan & Margolin, in press; Piotrkowski, Rapoport, & Rapoport, 1987); (b) the
association between parenting and marital relationships, and the influence of children on
marital relationships, has been well-documented (Brody, Pellegrini, & Sigel, 1986;
Easterbrooks & Emde, 1988; Renne, 1970); and (c) identifying the short-term effects of
stressors on family interactions may be crucial for the understanding of larger, long-term
effects (Caspi et al., 1987; Repetti, 1989).
Sociological theories of work and family have influenced psychological research into
the work-family interface. Such theories posit that the establishment of industrially-based
societies, where factories were the major source of employment, led to the dissolution of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
bond between work and family (Rapoport & Rapoport, 1965; Ridley, 1973). Researchers
thus began to regard work and family as isolated areas of interest, and propagated the “myth
of separate worlds” (Kanter, 1977). The predominant early tradition in social science theory,
then, assumed that work and family were disconnected, but equal, institutions.
Role theory provided the implicit conceptual framework for many early studies of the
work-family interface (see Bates, 1956; Ridley, 1973). This theory holds that the dominant
role for the male in industrial societies is that of the breadwinner, while roles such as wife’s
lover and children’s playmate are secondary. The latter roles may thus be more flexible, and
may be adjusted to fit the demands of the dominant role. Researchers thus identified the
workplace, both implicitly and explicitly, as the primary stressor. They viewed the home,
the wife’s domain, as a sanctuary from the rigors of the workday, a benign environment
where one could recuperate from work stress (Baruch, Biener, & Barnett, 1987).
Additionally, early research on work stress treated employment as a dichotomous
variable; studies failed to include occupational conditions, such as conflict between
coworkers, job satisfaction, and workload, as variables. These studies also focused
exclusively on men, since in traditional marriages they were the sole breadwinners.
Since the early 1960s, empirical studies have corrected some oversights of the early
investigators. The results of these studies have also countered the assumption that work and
family spheres are orthogonal. Investigators (e.g., Blood & Wolfe, 1960) inspired these
studies by addressing questions of how the husband’s occupational status affected family
variables such as marital satisfaction. Much research has since focused on such questions
as whether husbands’ job satisfaction affects wives’ marital satisfaction, and how husbands’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
work schedule affects their performance of family-related roles (e.g., Barling, 1984; Clark,
Nye, & Gecas, 1978; Piotrkowski, 1979).
Increasing numbers of women began to enter the labor market in the middle of this
century. This was a major social change and led researchers to consider more closely the
“myth of separate worlds.” Studies noted the progressive tendency for both members of the
marital dyad to adopt the roles of worker, parent, and spouse simultaneously. The emergence
of “two physically separated and socially distinct work roles for women: one inside and one
outside the home” (Piotrkowski & Repetti, 1984, p. 99) led clinicians and researchers to
focus on multiple role stress and role overload in employed wives (e.g., Baruch et al., 1987;
Burke & Weir, 1976; Ladewig & White, 1984; Pleck, 1985; Ridley, 1973; Yogev, 1986).
A focus on these issues was consistent with the social climate of the time, when many
psychologists viewed wives’ employment as incompatible with their traditional family role
(Repetti, 1987). Also, they assumed a child’s bond to his mother was most critical for his
emotional adjustment and development. So, the effects of wives’ employment began to
receive more attention than did those of husbands’ employment.
A major research paradigm in studies of the work-family interface at this time was
that of interrole conflict. This paradigm is based upon the “scarcity hypothesis” (Baruch et
al., 1987), which assumed that individuals have limited supplies of physical, mental, and
emotional resources. By this model, individuals, especially women, who accepted multiple
roles would find work and family spheres competing for their personal time and energy
resources. If one sphere depleted the resources, fewer would be available for use within the
other system, leading to role overload (Loscocco & Roschelle, 1991; Repetti, 1987). For
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
example, researchers theorized that if occupational stressors were present on a daily basis,
employees’ behavior at home would change because work-related problems and needs would
preoccupy them. Thus, as employees became “more withdrawn, less empathic, and less
effective and reliable family members” (Repetti, 1987, p. 121), family interactions would
change.
The enhancement hypothesis (see Baruch et al., 1987; Marks, 1977; Loscocco &
Roschelle, 1991; Piotrkowski & Repetti, 1984; Thoits, 1983) provided a competing model
to the scarcity hypothesis. This theory argues that multiple roles serve a protective function:
the more roles one takes on, the more status one gains, and the more one can assign the less
desirable aspects of certain roles to others. Accordingly, multiple roles are not necessarily
detrimental to either employees or to their families; some empirical studies (e.g., Burke and
Weir, 1976) have found that employment may, in fact, mitigate the psychological effects of
marital stress.
In contrast to many previous studies, this paper studies the effects of multiple roles
in dual-earner families, not merely dual-earner marriages. Thus, this study is unique within
the literature in that it includes the parent-child relationship as a variable.
Why do dual-eamer families merit special consideration? It appears that such
families produce different types of family relationships than do traditional single-eamer
families (Piotrkowski & Repetti, 1984). Role theory can also define the manner in which
married couples divide labor within their households. In traditional single-eamer families,
wives’ primary resource is time in the home, while their husbands’ primary resource is the
ability to gain employment outside the home. Wives thus perform more family work (i.e.,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
childcare and housework) than do their husbands (Blood & Wolfe, 1960). Thus, in
traditional couples, wives exchanged family work for the financial and emotional security
provided by their husbands; the differentiation and separation between the roles of worker,
parent, and spouse were clear (Pleck, 1985).
Dual-earner marriages blur this differentiation. Literature reviews in the mid-1960s
and 1970s noted that dual-eamer couples did not fit neatly into this theoretical framework
(Malmaud, 1980; Pleck, 1985). Rapoport and Rapoport (1965) and Burke and Weir (1976)
identified certain stresses, such role overload and interrole conflict, as more intense in dual-
eamer families than in traditional single-eamer families. These researchers thus
hypothesized that the division of family work in dual-eamer couples is not equitable:
employed wives continued to accept a larger share of the household labor than did their
husbands.
By this model, marital stress in dual-eamer families could arise from several sources
unique to the structure of such families: first, some studies (e.g., Pleck, 1985) suggested that
wives’ role overload has negative consequences for their emotional well-being. Second, an
unequal division of household labor could dissatisfy wives, and they might then insist their
husbands to do more family work. Third, husbands may resist their wives having potentially
more successful careers. Fourth, couples may struggle with time management in balancing
career and family (Malmaud, 1980). Finally, equity theory, social psychological principle,
states that an inequitable relationship (i.e., where one person gives more than they receive)
might dissatisfy both partners (Brown, 1986). On these bases, then, dual-eamer couples may
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
be exposed to different sources of stress and more marital dissatisfaction than are single-
eamer couples (Renne, 1970).
By the 1980s, husbands in dual-eamer marriages had significantly increased the time
they spent on housework and parenting tasks (Pleck, 1985). In dual-eamer families, time
spent on housework and direct interaction with children moved toward equity. Nonetheless,
wives in dual-eamer families still spent more time being available to their children
(performing what Pleck, 1985, p. 49 calls “the broader range of childcare activities”). So,
although role overload theory in its strictest form no longer applies, it may still be true that
employed wives encounter more multiple role stress than do their husbands (Malmaud, 1980;
Pleck, 1985). This, in turn, may be associated with marital relationship stress.
Clearly, dual-eamer families are different from single-eamer families in both their
structure and in the processes involved in their interactions. Because of this, and because
dual-eamer marriages are now the norm in the United States (Chan & Margolin, in press;
Piotrkowski et al., 1987), an examination of the work-family relationship within dual-eamer
families is necessary both for theoretical reasons and to satisfy statistical concerns about
generalizability. Nonetheless, few studies in this field have concentrated solely on dual-
eamer families. Some (e.g., Piotrkowski, 1979) excluded dual-earners to preserve sample
homogeneity, while others (e.g., Crouter et al., 1989; Repetti, 1989) did not make clear the
distinction between dual- and single-eamer families.
The factors, mechanisms, and processes that connect work and family systems are
possibly the most important aspects of research in this field. McKinley (1964) was probably
the first investigator to suggest work has an impact on daily patterns of family interaction.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
This researcher investigated the parenting style of male blue-collar workers, and linked a
perceived lack of autonomy at work to the use of strict disciplinary tactics at home.
Rapoport and Rapoport (1965) suggested patterns of family and work life affected
each other, that is, that family life may make demands that the occupational role must adapt
to, in the same way as the employee’s job may make demands on the family. Several
researchers (e.g., Kanter, 1977; Staines, 1980) used this suggestion as a touchstone. Most
studies focused on the impact of husbands’ jobs on the family (e.g., Chan & Margolin, in
press). A few, however, considered the relationship between wives’ employment status and
its effects on the marital interaction. Results of these latter studies are equivocal: some found
that the effects are likely to be negative in working women (e.g., Bronfenbrenner & Crouter,
1982; Kessler & McRae, 1982; Ridley, 1973), while others found protective effects of wives’
employment (e.g., Burke & Weir, 1976; Loscocco & Roschelle, 1991; Piotrkowski &
Repetti, 1984).
Overall, these studies generated little empirical data and no consistent, integrative
theories for the analysis of the work-family relationship. Most significantly, they did not
clarify the extent to which family affects work, and provided little insight into the processes
whereby family might influence work (Crouter, 1984). Methodological weaknesses were
the primary reason for these shortcomings. The following list highlights the most severe of
these weaknesses.
First, researchers conducted studies from a descriptive, qualitative perspective;
longitudinal studies have not replicated many of these findings and conclusions. Second,
studies had a unidirectional focus on the effects of work on the marital relationship, and did
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
not explore the effects that marriage can have on work. Finally, the fact that these studies
ignored dual-eamer couples represents a serious compromise of generalizability (Chan &
Margolin, in press; Hoffman & Hoffman, 1985).
Refinement of methodology and development of psychological theory were
necessary for research in this field to advance. Piotrkowski’s (1979) inquiry into mediating
processes in the work-family relationship provided much impetus for subsequent theoretical
advances. Her pioneering hypothesis stated that three discrete psychological processes can
capture the paid worker’s psychological reactions to work, and the manner in which these
reactions subsequently penetrate the home environment: (a) Positive carry-over: the
employee has an emotionally satisfying day at work and returns home with high emotional
energy and an availability for family interactions; (b) negative carry-over: the employee
experiences stress at work and requires other family members to expend their resources in
helping manage these stresses at home; and (c) energy deficit: the employee is physically
and/or mentally exhausted by the day at work and is subsequently less involved in family
interactions than normal. Piotrkowski (1979) was thus the first researcher to identify
psychological processes that link work and family.
Other researchers built on these ideas. Since the early 1980s much of the literature
in this field has focused primarily on a bidirectional (i.e., work-to-home and home-to-work)
spillover model (e.g., Barling & Rosenbaum, 1986; Burke, 1982; Crouter, 1984; Piotrkowski
& Crits-Cristoph, 1982; Repetti, 1987). Investigators established a consistently significant,
if modest, association between work-induced emotional states and moods, and similar states
and moods in the home. For example, husbands’ exposure to work stressors was associated
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
with negative family interactions and reduced responsiveness to, and involvement with, other
family members, (Crouter, 1984; Repetti, 1987). Conversely, wives’ job mood was
positively associated with their home mood; their job satisfaction was positively correlated
with marital satisfaction, home mood and satisfaction with family relationships (Piotrkowski
& Crits-Cristoph, 1982).
These studies represented an advance over their predecessors because they clearly
focused on the spillover link between work and family. They also gathered considerable
empirical evidence. The research designs they employed, however, did not succeed in
correcting some methodological failings of earlier works. Many findings were based on case
studies and small samples— for example, Piotrkowski (1979) worked with only 13 families.
This raises questions about the generalizability of the conclusions. In addition, the continued
employment of cross-sectional research designs, semistructured interviews, and naturalistic
observational techniques could not resolve more serious shortcomings in methodology.
Two general categories, inflated correlations and inconclusive causal priority, define
these shortcomings (Bolger, DeLongis, Kessler, & Wethington, 1989; Repetti, 1989).
Inflated correlations can occur because of a heavy dependence on self-report assessments by
only one person (an investigator or the working spouse) of conditions both at work and at
home. Piotrkowski and Crits-Cristoph (1982), for example, asked wives to list the mood
states they had experienced at work and at home over the previous two days. While the
researchers were aware of the problems of accurate recall in such a design, they did not
correct for the possible influence of current mood on recall and memory. They also failed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
to account for potentially inflated correlations arising from respondent biases in perceiving
the same mood states across different situations (Repetti, 1989).
Inconclusive causal priority refers to the uncertainty in establishing the direction—
work-to-home or home-to-work— in which stress spillover primarily occurs. This is a major
methodological problem resulting from the use of cross-sectional designs. Although
investigators using such designs acknowledged that work-family relations are most likely
circular, they could not empirically distinguish causal processes without making unfounded
inferential assumptions.
This methodological shortcoming is the root of several concerns that have surrounded
this field of research (Bolger et al., 1989). The first is stated above: cross-sectional survey
and interview data cannot directly document the mediating processes involved in the work-
family relationship. Investigators cannot interpret significant positive correlations as
showing that stress does spill over from work to home: even without a true stress spillover,
a significant correlation could also indicate an unmeasured cause of stress (a “third variable”)
common to work and home environments.
Second, even if we assume qualitative data can directly document the processes, they
leave unanswered the question of the direction in which spillover occurs: is it primarily from
home to work, work to home, or in equal quantities in both directions? Most early research
assumed that spillover was mainly from work to home, but empirical results suggest that
home-to-work spillover may be equally plausible (Staines, 1980).
Third, qualitative studies did not address the crossover effect, that is, whether and
how chronic work stresses affect the worker’s spouse, and what patterns of family adaptation
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
do they set in motion? Finally, qualitative studies exploring sex differences in stress
spillover obtained equivocal results. For example, Pleck (1977) found a more pronounced
work-to-home spillover in husbands, while home-to-work spillover was stronger in wives.
Crouter (1984) suggested, based on participants’ subjective perceptions, that home-to-work
spillover was stronger in wives. Pleck and Staines (1985), in a cross-sectional survey study,
found that associations between work variables and spousal functioning were as strong in
husbands as they were in wives.
Recent studies in this field (e.g., Bolger et al., 1989; Caspi et al., 1987; Crouter et al.,
1989; Doumas, 1993; Repetti, 1989) have responded to the imbalance between well-defined
theory and poor research design by incorporating several methodological innovations. These
include four strategies: (a) Testing specific hypotheses, based firmly on the theory developed
by qualitative studies, using quantitative methodology; (b) obtaining data from both
husbands and wives; (c) utilizing within-subjects designs where earlier investigators had
employed between-subjects designs; and (d) using repeated daily measures in longitudinal
designs to examine sequential and possible causal relationships in the spillover and crossover
processes.
Within-subjects designs make more sense in the present context, since daily
variability in experience of the work environment and the family are unique to each person.
Also, daily stressors seem to make an independent contribution to psychological well-being,
and may mediate the effects of major fife stressors (Caspi et al., 1987). Identifying the short
term effects of daily stressors on family interactions is as important as comprehending the
impact of major life events and longer-term stressors (Bolger et al., 1987; Repetti, 1989).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
Crouter et al. (1989) and Repetti (1989) conducted studies that employed these
methodological characteristics. Crouter and colleagues found that mens’ self-reported high
stress level after work was associated with increases in their wives’ reports of negative
marital interactions later that evening. Repetti (1989) found positive associations between
male air traffic controllers’ work stress and their social withdrawal and expressions of anger
at home later that day.
The latter study also found that the degree of social support shown by their wives
affected husbands’ withdrawal and anger. More social support led to greater withdrawal and
fewer expressions of anger. Therefore, supportive spouses may “buffer” the effects of daily
stressors by allowing the worker to withdraw and return to a baseline emotional and
physiological state. Loscocco and Roschelle (1991) concluded, after a literature review, that
there is some evidence that social support diminishes negative effects of work stress, even
though there is no direct indication that social support moderates the relationship between
job stress and mental health. They suggest that the key variables may be the form of
buffering (i.e., the type of social support) and the person who provides the social support.
Bolger et al. (1989) conducted the most comprehensive longitudinal study of the
work-family interface yet published. Based on previous qualitative and quantitative research,
they introduced the idea of “stress contagion,” which describes the manner in which multiple
role functioning generates daily stress. It involves two components: (a) Stress spillover: the
stresses experienced by an individual in either the work or the home domain lead to stresses
for that individual in the other domain, and, (b) stress crossover: the stresses experienced by
one’s spouse at work lead to stresses for oneself at home.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
To study these effects, Bolger et al. (1989) employed a longitudinal diary study of
166 married couples over 42 consecutive days. Using a time-series regression analysis, they
found that husbands were more likely than wives to bring home stresses into the workplace.
Stress contagion from work to home was evident for both husbands and wives. These
investigators’ examination of the contagion of work stress into the home uncovered a process
of dyadic adjustment: one spouse (usually the wife) modified (i.e., increased) her housework
efforts to compensate for the work stresses evident in her partner. This provides support for
the buffering effect noted by Repetti (1989) and Loscocco and Roschelle (1991), and
suggests significant sex differences in role stress. Clearly, then, recent studies in this field
have given prominence to the crossover hypothesis, sex differences, and the phenomenon of
buffering or social support.
Similar to Bolger et al. (1987) and Doumas (1993) studies, the present study
examines the daily variability and the sequential, bidirectional nature of the spillover and
crossover processes. It also attempts to resolve the “third variable” problem described above.
The present study is the first of its kind to focus exclusively on dual-eamer families. It
recognizes their differences from single-eamer families and the unique status that they must
thus possess in the study of work-family linkages (Malmaud, 1980; Pleck, 1985; Yogev,
1986).
Furthermore, the present study examines the parent-child relationship as a variable
in stress contagion models. Studies have identified this relationship as an important factor
in marital satisfaction (see Brody et al., 1986; Easterbrooks & Emde, 1988; Renne, 1970).
A major omission of the research tradition in the work-family interface is that most
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
researchers have not considered the influence of children on the family system; some
samples described as consisting of dual career families are more accurately characterized as
consisting of dual career couples (Piotrkowski & Repetti, 1984).
Recent research has provided evidence that including children in any examination of
the work-family interface is essential. For example, Crouter (1984) reports that, in her
sample, patterns of spillover systematically related to the presence or absence of children.
Specifically, she found that mothers reported greater difficulty than fathers in terms of
negative spillover, but that there was little difference between childless spouses in this
regard. Repetti (1987) found that the presence or absence and age of children influence the
degree to which job stress affects families. Also, it appears wives may increase their
involvement with their children when their husbands experience high work stress days
(Pleck, 1985; Repetti, 1989). Finally, Baruch and colleagues’ (1987) assert that employment
may buffer the effects of marital stress for wives, but may increase their vulnerability to the
effects of parental stress.
This study attempted to disconfirm the notion that there is no relationship between
daily work stress and daily family interactions. To this end, it used spillover and crossover
models to examine intimacy, withdrawal, and conflict within the marital relationship, and
affection, involvement, and conflict within the parent-child relationship. The following
hypotheses were tested:
Spillover Effects
Hypothesis One. An individual’s daily work stress will be predictive of his or her
family interactions the same day. For example, high work stress predicts to: (a) Low
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
intimacy, high conflict, and/or high withdrawal in the marital relationship the same day for
that individual, and (b) low affection, high conflict, and/or low involvement in the parent-
child relationship the same day for that individual.
Hypothesis Two. An individual’s daily family interactions will be predictive of his
or her work stress the following day. For example: (a) High intimacy, low conflict, and/or
high withdrawal in the marital relationship predict to low daily work stress the following day
for that individual, and (b) high affection, low conflict, and/or high involvement in the
parent-child relationship predict to low daily work stress the following day for that
individual.
Crossover Effects
Hypothesis Three. An individual’s daily work stress will be predictive of his or her
spouse’s family interactions the same day. For example, high work stress predicts to: (a)
Low intimacy, high conflict, and/or high withdrawal in the marital relationship for the spouse
on the same day, and (b) low affection, high conflict, and/or low involvement in the parent-
child relationship for the spouse on the same day.
Hypothesis Four. An individual’s daily family interactions will be predictive of his
or her spouse’s work stress the following day. For example: (a) High intimacy, low conflict,
and/or high withdrawal in the marital relationship predict to low work stress for the spouse
the following day, and (b) high affection, low conflict, and/or high involvement in the
parent-child relationship predict to low work stress for the spouse the following day.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
Method
Participants
The forty-four families who participated were a subgroup of a sample of 180 intact
families who formed part of a larger study (see Margolin & John, 1988). This latter study
recruited families through direct mailings and public announcements in the Los Angeles
area. Participating families met the following criteria: (a) the family contained a child aged
four to five years old inclusive (the “target child”), (b) both adults were the biological parents
of the target child, (c) the target child was participating in a preschool, nursery school, or
kindergarten program outside the home, and (d) both parents read and spoke English.
Families received 90 dollars for participating in a three-hour laboratory session. For
an additional 20 dollars, they filled out the Adult Home Data Questionnaire (see Appendix
A) every day for the two weeks following the laboratory session. They also received a five-
dollar bonus for promptly returned and error-free Adult Home Data Questionnaires.
Dual-eamer families who accepted the above option comprise this study’s sample.
The present study defines dual-eamer families as those in which each partner worked in the
labor market for 40 hours or more over the two-week period of data collection. Establishing
a cutoff of 20 hours per week per partner as the minimum for dual-eamer family status has
precedence in the field (see Locksley, 1980). The final sample of 44 families was selected
by calculating each participant’s 14-day total for Item 69 (“How many paid hours did you
work today?”) of the Adult Home Data Questionnaire.
Table 1 shows some demographic characteristics from the sample. Eighteen (40.9%)
of the target children were male, while 26 (59.1%) were female. The ethnic composition of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
the sample was as follows: among husbands, 30 (68.2%) were Caucasian, 8 (18.2%) were
African-American, 4 (9.1%) were Hispanic, and 2 (4.5%) described themselves as “Other”;
among wives, 31 (70.5%) were Caucasian, 5 (11.4%) were African-American, 5 (11.4%)
were Hispanic, 2 (4.5%) were Asian or Pacific Islander, and 1 woman (2.3%) described
herself as “Other.” Among couples, 28 (63.6%) were Caucasian, 10 (22.7%) were biracial,
4 (9.1%) were African-American, 1 (2.3%) was Hispanic; one couple described themselves
as “European.”
Table 1. Means and Standard Deviations for Selected Couple.
Spouse, and Target Child Variables
Variable
M
SD
Couple Characteristics
Family income (dollars/month) 5,031 2,442
Number of children 2.1 0.9
Number of years married 9.4 4.4
Husband Characteristics
Age (years) 36.5 5.8
Years of education 15.0 2.1
Hours worked per week 48.0 15.6
Wife Characteristics
Age (years) 34.1 4.7
Years of education 15.0 2.0
Hours worked per week 38.5 11.9
Target Child Characteristics
Age (years) 4.5 0.5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
Procedures
Husband and wife self-reports of daily events from the Adult Home Data
Questionnaire provided the data for this study. These self-reports followed a three-hour
laboratory session in which parents completed packets of questionnaires while an
experimenter verbally administered questionnaires to their child. At the conclusion of this
stage, experimenters asked each parent to fill out the first Adult Home Data Questionnaire;
they then trained the parents on procedures for filling out the questionnaire on a daily basis.
To reduce errors, families were alerted to seven common problems that arise when
completing the questionnaires: (a) not filling out “Day of Activity” and “Day Reported”
correctly, (b) skipping question 54 and question 56, (c) forgetting to circle “no” responses
on questions 76-87, (d) skipping questions, (e) circling more than one response, (f) forgetting
rules regarding days of no contact with spouse or child, and (g) leaving question 88 blank.
Each participating family received a sheet explaining these problems in detail (see Appendix
B).
Participants ideally filled out the questionnaires on their own every evening about a
half-hour before bedtime. If this was impossible, they filled it out when they could,
indicating the date on which they completed it at the top of the form. To preserve
confidentiality, they used stickers to seal the questionnaires after completion. Each
participant received 13 questionnaires, 13 sealing stickers, and 13 stamped addressed
envelopes.
A five-dollar bonus was offered to encourage same-day reporting and to reduce
errors. Participants received the five dollars if they exhibited perfect reporting (i.e., if the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
postmarks on envelopes showed the questionnaires had been mailed every day, and if each
questionnaire contained fewer than five errors). Seventy-six percent (938 of 1232) of the
questionnaires were returned on time. If errors were committed, experimenters telephoned
families to implement corrections, to acquire any missing data, and to clarify unclear
responses (Margolin and John, 1988).
The Human Subjects Committee approved all procedures.
Measures
Adult Home Data Questionnaire. The Adult Home Data Questionnaire (Margolin,
1989; Appendix A) is an 88-item self-report questionnaire designed to gain information on
daily events within and outside the family. Doumas (1993) used an earlier version of this
questionnaire to examine the relationship between marital interaction, stress and health
behaviors. In the present study, participants completed the questionnaire on a daily basis for
14 days.
Indices for the constructs of work stress, marital interaction (consisting of conflict,
withdrawal, and intimacy), and parent-child interaction (consisting of conflict, involvement,
and affection) in each parent were created by submitting selected subgroups of items to
confirmatory factor analyses with varimax rotation. Items with a loading of greater than .60
were retained as measures of the factors in each case.
As expected, several items combined to create indices of work stress, parent-child
affection, parent-child conflict, parent-child involvement, and marital intimacy. Contrary
to expectation, the marital conflict items yielded two distinct factors (verbal conflict between
partners and physical conflict between partners). The marital withdrawal items also yielded
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
two distinct factors (withdrawal from partner, and perceived withdrawal by partner). Each
variable is defined in more detail below.
Work Stress
The present study defines this variable as the stress experienced by an individual due
to work. This stress may be due to workload, conflict with co-workers, and amount of time
spent at work. A work stress index was obtained by summing the scores of items 1,2,4, 5,
6,12,15,18, and 69 of the Adult Home Data Questionnaire. These items are listed below.
1. My stress level at work today was: a) high b) moderate c) low d) none
2. My work load today was: a) high b) average c) low d) none
4 .1 was irritated/annoyed by someone at work today:
a) a lot b) a little c) not at all
5. Someone at work was irritated/annoyed with me:
a) a lot b) a little c) not at all
6 .1 still felt stress or pressure from work when I was home with my family:
a) a lot b) some c) a little d) none
12. Today the stress I felt from other than my family (work, friends,
activities) was: a) high b) moderate c) low d) none
15. Today I felt stressed by decisions: a) a lot b) some c) a little d) none
18. Today I felt stressed by tasks needing my attention:
a) a lot b) some c) a little d) none
69. How many paid hours did you work today?
a) none b) <4 hrs c) 4-8 hrs d) 8-12 hrs e) > 12 hrs
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
Marital Interaction
Intimacy. The present study defines this variable as both physical (kissing, hugging,
sexual intercourse) and intellectual (having a good conversation) expressions of affection
between partners. An index was obtained by summing the scores of items 42, 43, and 73
(where a response of “a” was scored as 5 and one of “b” as 1) of the Adult Home Data
Questionnaire. These items are listed below.
42. My partner and I kissed and hugged each other today:
a) not at all b) a little c) some d) quite a bit e) a lot
43. My partner and I had a good conversation today:
a) not at all b) a little c) some d) quite a bit e) a lot
73. Did you have sexual intercourse within the last 24 hours? a) yes b) no
Conflict. Following the factor analysis discussed above, the present study defines
this variable on two levels: physical conflict (e.g., hitting, pushing, shoving) and verbal
conflict (verbal expressions of anger). An index of physical conflict was obtained by
summing each individual’s scores on items 40 and 41 of the Adult Home Data
Questionnaire. These items are listed below.
40. Today I hit, pushed, or shoved my partner:
a) not at all b) a little c) some d) quite a bit e) a lot
41. Today my partner hit, pushed, or shoved me:
a) not at all b) a little c) some d) quite a bit e) a lot
An index of verbal conflict was obtained by summing each individual’s scores on
items 36 and 37 of the Adult Home Data Questionnaire. These items are listed below.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
36. Today I expressed anger or irritation at my partner:
a) none b) a little c) some d) quite a bit e) a lot
37. Today my partner expressed anger or irritation at me:
a) none b) a little c) some d) quite a bit e) a lot
Withdrawal. Following the factor analysis discussed above, the present study defines
this variable on two levels: withdrawal by partner (a report by the respondent that his or her
spouse withdrew from interaction) and withdrawal by partner (a report by the respondent that
he or she withdrew from interaction with his or her spouse). An index of withdrawal from
partner was obtained by summing each individual’s scores on items 47 and 48 of the Adult
Home Data Questionnaire. These items are listed below.
47.1 ignored my partner’s wishes or needs:
a) not at all b) a little c) some d) quite a bit e) a lot
48.1 took my partner’s feelings lightly:
a) not at all b) a little c) some d) quite a bit e) a lot
An index of withdrawal by partner was obtained by summing each individual’s scores
on items 38, 39,46, and 49 of the Adult Home Data Questionnaire. These items are listed
below.
38.1 felt distant or withdrawn from my partner:
a) not at all b) a little c) some d) quite a bit e) a lot
39. My partner seemed distant or withdrawn from me:
a) not at all b) a little c) some d) quite a bit e) a lot
46. My partner ignored my wishes or needs:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
a) not at ail b) a little c) some d) quite a bit e) a lot
49. My partner took my feelings lightly:
a) not at all b) a little c) some d) quite a bit e) a lot
Parent-Child Interaction
Affection. The present study defines this variable as the number of physical acts of
hugging, kissing or other displays of affection between parent and child. An index of parent-
child affection was obtained by summing the scores of items 22 and 32 of the Adult Home
Data Questionnaire. These items are listed below.
22. How much love or affection did you show your child today?
a) none b) a little c) some d) quite a bit e) a lot
32. How much did you hug or kiss your child today?
a) none b) a little c) some d) quite a bit e) a lot
Conflict. The present study defines this variable as how much anger and yelling is
present in the daily interaction, how often the parent either threatened to or did punish the
child, and how often the parent was irritated by the child. An index of parent-child conflict
was obtained by summing the scores of items 23,24,25,28, 29 and 30 of the Adult Home
Data Questionnaire. These items are listed below.
23. How angry were you at your child today?
a) none b) a little c) some d) quite a bit e) a lot
24. How angry was your child at you today?
a) none b) a little c) some d) quite a bit e) a lot
25. How much did you punish your child today?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
a) none b) a little c) some d) quite a bit e) a lot
28. How irritated were you with your child today?
a) none b) a little c) some d) quite a bit e) a lot
29. How often did you warn your child she or he might be punished?
a) none b) a little c) some d) quite a bit e) a lot
30. How much did you yell at your child today?
a) none b) a little c) some d) quite a bit e) a lot
Involvement. The present study defines this variable as the number of different
activities the parent and child engaged in together. The literature generally defines parent-
child involvement more broadly, primarily focusing on the interest parents show in all
aspects of their child’s learning. Studies in this area have, however, also confirmed that
effective parents engage in more activities with their children than do less confident parents:
simply doing things with one’s children seems to have a significant influence on their
functioning (Swick, 1988). By including this dimension of the parent-child relationship, the
present study can examine the effects work stress may have on parenting.
An index of parent-child involvement was obtained by summing items 76-87 of the
Adult Home Data Questionnaire. On these items, parents were asked if they had engaged
in any of a series of activities with the target child on that day. If the parent responded
“YES” to any of the items, he or she was given a score of 1. A response of “NO” was given
a score of 0. These items are listed below.
76) Playing 82) Discussing child’s day
77) Shopping 83) Cooking, chores
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
78) Doing homework 84) Sports, board games
79) Reading 85) Going out to an event
80) Taking a walk 86) Eating together
81) Dance, music, art 87) Watching TV or VCR
Results
Data Analysis
Since this study focused on within-person variation in the work-family relationship,
individual effects in the data were removed using z-scores. The first step in the data analysis
was therefore to calculate each participant’s daily score for each variable. Once this was
completed, each participant had 14 different daily scores and a mean score for each variable.
Each daily score on each variable was then subtracted from the mean score on that variable
for each participant. The results of these calculations were standardized, thus laying the
foundation for a within-subjects data analysis.
An analysis based on linear regression techniques was appropriate because the
variables in this study were continuously distributed. Performing an ordinary linear
regression, however, is erroneous when faced with data of the present form. Thus, the data
analysis was performed by using a pooled time series autoregression (Bolger et al., 1989;
Caspi et al., 1987; Doumas, 1993; Gottman, 1981; Neter, Wasserman, & Kutner, 1983;
Ostrom, 1990).
Time series data are problematic for regression analyses because the statistical
assumptions underlying such analyses are frequently invalid for such data (see Caspi et al.,
1987, for a detailed discussion; Ostrom, 1990; Gregson, 1983). One assumption of an
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
ordinary regression analysis is that residuals or error terms from different measurement times
are orthogonal or uncorrelated. Autocorrelation, however, is typical of time series data,
since most variables social scientists measure at any given time have approximately the same
value as at their previous time of measurement. Thus, time series regression frequently
violates the assumption of uncorrelated errors.
Caspi et al. (1987) propose two possible explanations for the presence of
autocorrelated errors in time series data. The first involves measurement error. If the same
inaccuracies in measurement occur during successive assessments, autocorrelated errors may
occur. The most common cause of autocorrelated errors, however, is the failure to include
important explanatory variables in the regression equation. Naturally, it is extremely
difficult, if not impossible, to include all important explanatory variables in the equation.
If some of the omitted variables, however, are themselves autocorrelated and follow a stable
course over time, they will influence the covariances among error terms, cause systematic
variation in the modeled behavior, and lead to correlations between temporally adjacent error
terms.
The problem of autocorrelation can have an important impact on the quality of the
inferences one can draw from empirical analyses. When a regression analysis violates the
assumption of uncorrelated error terms, the significance levels and goodness-of-fit statistics
reported by ordinary linear regression procedures are unreliable, and the R2 value does not
accurately summarize the explanatory power of the independent variables. It should thus be
clear that autocorrelation in a regression model does not bias the coefficients: their expected
value is still found in the population. Autocorrelation does, however bias the standard errors
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
of the coefficients downwards, and so estimates of the influence of independent variables on
the dependent variable are more precise than is actually true. In the worst case, this could
lead one to falsely conclude that variables in the sample are related to each other when in fact
they are unrelated in the population.
Other important factors make an ordinary regression analysis inappropriate for use
with time series data. One of these is that most series exhibit either an upward or downward
trend. Therefore, any two trending time series will correlate simply because of the trends,
no matter whether a causal relationship is present or not. Ordinary regression with serially
correlated time series may thus be unreliable: an increasing trend will likely continue to
increase, but that does not mean one can or should use another trend to predict the first
trend’s continued increase. When using a regression analysis, the aim is to obtain estimates
of the linear relationship between series apart from accidental similarities resulting from
autocorrelation.
This study is concerned both with individual differences and intraindividual
variability over time. To examine both questions in the same model, the analysis was
performed using pooled data. The use of pooled data allows one to incorporate individual
and daily level variables in a single model (Caspi et al., 1987; Judge, Griffiths, Hill,
Lutkepohl, & Lee, 1985). By pooling the data of all the participants into a single matrix, one
obtains a vast increase in degrees of freedom over individual regression equations.
Parameters can thus be estimated with greater efficiency.
The pooling method used here resulted in a data matrix with N x T rows, where N is
the number of cases (44) and I is the number of repeated measurements per case (14). Rows
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
of the matrix indexed days, and columns indexed variables. The first block of 14 rows in the
matrix contained data concerning the first family, the second block contained data
concerning the second family, and so on.
The issue of autocorrelation is as important for pooled data as it is for a single set of
repeated measures. This analysis assumed there was a single value for the autocorrelative
term (i.e., that this term did not vary significantly between individual participants). By this
assumption, estimating an average autocorrelative factor adequately captures the
autoregressive process. Thus, with some care and with the use of pooled time series
regression analysis, one can overcome the autocorrelation problem and other difficult issues
related to time series data.
The analysis examined spillover and crossover effects of work stress to family
interactions in the same equation, which followed this basic model:
FIit = f[FIi(t_ i) + WSit + WSj, + Controls] (1)
This model assumes family interaction for individual I on day t (FIit ) is a function of family
interaction for the same individual on the previous day (FI^.,)), that individual’s work stress
on the same day (Wsi t, the spillover effect), the spouse’s work stress on the same day (Wsjt,
the crossover effect), plus control variables.
The analysis examined spillover and crossover effects of family interactions to work
stress in the same equation, which followed this basic model:
WSit = ftWSjft.,, + FIt> 1 ) + FIj(t.,) + Controls] (2)
This model assumes work stress for individual I on day t (WSit) is a function of work stress
for the same individual on the previous day (WS^.,)), that individual’s family interaction on
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
the previous day (FIl ( t.I), the spillover effect), the spouse’s work stress on the same day (Fij(t_
the crossover effect), plus control variables. Note that each model incorporates a time lag
effect, indicated by the subscript t-1, which accounts for the daily prediction of family
interactions from work stress and vice-versa.
Previous studies using this type of analysis (e.g., Bolger et al., 1989; Doumas, 1993)
included day of the week and participant’s time in the study as control variables.
Researchers in those previous studies expected that, first, participants would experience low
work stress and increased family interactions on weekends, and, second, that participants
closer to the end of the study period would, due to fatigue, fill out the questionnaires out
differently. It was, however, apparent from perusal of the present data that most participants
in this study worked on weekends, and did not feel significantly less work stress at that time.
Also, this study had a limited data collection period (only 14 days, compared with 42 days
in both Doumas, 1993, and Bolger et al., 1989), and so fatigue was not a factor.
Control variables in the present models thus included only a standard regression
constant. As noted previously, the models implicitly control for temporally stable variables
that vary only across individuals: z-scores, obtained by standardizing each participant’s score
on each variable’s index, removed individual effects and between-person variation and were
used in the analysis.
Model (1) was broken down into eight sub-models, five of which estimate the
dimensions of marital interaction considered here, and the remainder of which estimate the
dimensions of the parent-child interaction considered here. For example, the following
model estimated marital verbal conflict (VC):
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
VCit = ftVCift.,, + WSit + WSjt + Controls] (3)
Analogous models estimated physical conflict, withdrawal from partner, withdrawal by
partner, intimacy, parent-child conflict, parent-child involvement, and parent-child affection.
Model (2) above estimated work stress for each participant.
The present data analysis used the Autoregression procedure in SPSS for Windows,
Release 6.0. This procedure overcomes the abovementioned difficulties in analyzing time
series data by assuming that adjacent error terms are autocorrelated. The Autoregression
procedure offers several different algorithms; the one used here, maximum likelihood, enters
variables separately and simultaneously into the regression equation and can tolerate missing
data in the series.
Descriptive Data
Tables 2-4 present means, medians, and standard deviations for the variables under
discussion for both husbands and wives. Wives’ mean daily raw score on each measure was
higher than that of husbands, with one exception: on the Withdrawal From Partner variable,
wives’ mean daily raw score was equal to husbands’. A series of oneway ANOVAs were
conducted on the mean scores. Wives’ daily work stress scores were significantly higher
than husbands’, F(l, 1230) = 9.55, g < .002. For marital relationship variables, wives scored
significantly higher on daily measures of Verbal Conflict, E(l, 1230) = 8.52, g < .003, and
Withdrawal by Partner, F(l, 1230) = 19.05, g < .000. For parent-child relationship variables,
wives scored significantly higher on daily measures of Affection, F(l, 1230) = 104.76, g <
.000, Conflict, E(l, 1230) = 27.78, g < .000, and Involvement, F(l, 1230) = 117.87, g < .000.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
Table 2. Descriptive Statistics for Participants’ Daily
Self-Reports of Work Stress
Variable*
M
Median SD
Husbands’ Report
Work Stress 17.04 17 5.30
Wives’ Report
Work Stress 17.99 17 5.43
Note. Values are derived from each participant’s raw score
on each variable for each day of the study.
“ Range of possible raw scores = 0-35.
These results appear to suggest that wives experienced more work stress, were more
sensitive to fluctuations in the marital relationship, and were more active in the parent-child
relationship. Such an interpretation may not be accurate, however. ANOVAs on the mean
scores may have detected significant differences between spouses because of the statistical
power provided by the large number of available degrees of freedom. Otherwise stated, the
significant effects detected by the ANOVAs in this sample may be statistical artifacts and
may not accurately reflect population effect sizes. The median raw daily scores seem to
support this argument: median scores for husbands and wives differed only for parent-child
relationship variables. Median scores on the other variables did not differ. A more plausible
interpretation, then, might be that wives were more active than husbands in the parent-child
relationship, but that there were no significant differences between spouses on work stress
and marital relationship variables.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
Table 3. Descriptive Statistics for Participants’ Daily Self-Reports
of the Marital Relationship
Variable
M
Median SD
Husbands’ Report
Intimacy1 1 6.84 6 2.82
Physical conflict6 2.23 2 .64
Verbal conflict 2.76 2 1.28
Withdrawal from
partner1 1
2.49 2 .95
Withdrawal by partner® 5.05 4 1.88
Wives’ Report
Intimacy2 6.91 6 2.87
Physical conflict6 2.24 2 .65
Verbal conflict® 2.98 2 1.39
Withdrawal from
partner1 1
2.49 2 .93
Withdrawal by partner® 5.62 4 2.67
Note. Values are derived from each participant’s raw score on each
variable for each day of the study.
“ Range of possible raw scores = 0-15. b Range of possible raw
scores = 0-10. “ Range of possible raw scores = 0-10. d Rangeof
possible raw scores = 0-20. 'Range of possible raw scores = 0-10.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
Table 4. Descriptive Statistics for Participants’ Daily
Self-Reports o f the Parent-Child Relationship
Variable
M
Median £D
Husbands’ Report
Affection1 6.70 7 2.01
Conflict6 8.14 7 2.68
Involvement0 3.56 4 2.05
Wives’ Report
Affection1 7.82 8 1.81
Conflict6 9.08 8 3.50
Involvement0 4.81 5 2.00
Note. Values are derived from each participant’s raw score
on each variable for each day of the study.
a Range of possible raw scores = 0-10. b Range of possible
raw scores = 0-30. c Range of possible raw scores = 0-12.
Preliminary Data
The data presented in Tables 5-9 have no direct bearing on the hypotheses under
discussion. The tables demonstrate, however, that the time series regression analysis used
in this paper can detect daily patterns in work stress and the marital and parent-child
relationship. The results shown in Table 5, for instance, clearly depict two patterns: first,
it appears the respondent’s work stress for one day significantly affects their work stress the
following day. Second, it appears work stress experienced by the respondent’s spouse on
one day has a significant positive association with the work stress experienced by the
respondent the following day. The observed effects are equally strong for men and women.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
The data presented in Tables 6 and 7 indicate participants perceive that intimacy on
one day has a significant positive association with intimacy the following day. Both spouses
report a similar pattern withdrawal by partner, and husbands report the same pattern for
verbal conflict. This suggests that both positive and negative interactions between partners
continue for more than one day at a time. An important caveat, to this statement, however,
is that both spouses report a significant negative association between physical conflict on one
day and physical conflict the following day. This suggests physical conflict between partners
is unlikely to persist for more than one day at a time.
for Work Stress Variables on One Dav Predicting Work Stress Outcomes
the Following Dav
Husbands Wives
Variable B SEE B SEB
Husbands’ Report
Work stress .19*** .04 .33*** .04
Wives’ Report
Work stress .34*** .04 .15*** .04
* * * £ < . 001.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 6, Day-to-dav Marital Relationship: Summary of Regression Analysis for
Marital Relationship Variables on One Day Predicting Husbands’ Experience of
Marital Interactions the Following Dav
Intimacy Physical Verbal Withdrawal Withdrawal
Conflict Conflict from Partner by Partner
Variable
Intimacy
B
SEB
Physical
Conflict
B
SEB
Verbal Conflict
B
SEB
Withdrawal
from Partner
B
SEB
Withdrawal
by Partner
B
SEB
.08*
.04
- . 11* *
.04
. 10*
.04
.05
.04
.16***
.04
* 1 2 < .05. **12 < .01. ***P <.001.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
Table 7. Day-to-day Marital Relationship: Summary o f Regression Analysis for
Marital Relationship Variables on One Dav Predicting Wives’ Experience of
Marital Interactions the Following Dav
Intimacy Physical
Conflict
Verbal
Conflict
Withdrawal
from Partner
Withdrawal
by Partner
Variable
Intimacy
B .09* — — — —
SEE .04 -- -- -- —
Physical
Conflict
B --
I
O
0 0
*
-- -- --
SEE
— .04 — - - - -
Verbal Conflict
B — - - .08 - - —
SEE
— — .04 — —
Withdrawal
from Partner
E
- - - - - - .04 —
SEE
— — — .04 —
Withdrawal
by Partner
E
— -- — -- .16***
SEE
-- — — — .04
*£<.05. ***£<.001.
The data presented in Tables 8 and 9 suggest participants perceive that affection
between parent and target child on one day has a significant positive association with
affection between the same two family members on the following day. As with the marital
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
relationship, this suggests that positive interactions in the parent-child relationship continue
for more than one day at a time. Husbands also report a significant negative association
between conflict with the target child on one day and conflict with that child the following
day. This suggests that negative interactions in the husband-child relationship do not persist
over several consecutive days.
Table 8. Day-to-day Parent-child Relationship: Summary of Regression Analysis for
Husband-child Relationship Variables on One Dav Predicting Husbands’ Parent-child
Relationship the Same Dav
Affection Conflict Involvement
Variable B SEB B SEB B SEB
Husbands’ Report
Affection .12* .04 -
Conflict -- - -.09* .04
Involvemen — — -- -- .07 .04
t
_ _
Table 9. Dav-to-dav Parent-child Relationship: Summary of Regression Analysis for
Wife-child Relationship Variables on One Dav Predicting Wives’ Parent-child
Relationship the Same Day
Affection Conflict Involvement
Variable B SEB B SEB B SEB
Wives’ Report
Affection .08* .04
Conflict — — .03 .04
Involvemen -- — -- — -.01 .04
t
*p < .05.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
Work-to-home Stress Contagion: The Marital Relationship
The data provide support for a prediction made by Hypothesis One. The results
shown in Tables 10 and 11 indicate a significant spillover effect from work stress to marital
intimacy: both husbands and wives report that their levels of work stress on one day were
negatively associated with the intimacy they experienced later that day.
The data also provide support for two predictions made by Hypothesis Three. First,
the results shown in Tables 10 and 11 indicate a significant crossover effect from one
spouse’s work stress to intimacy experienced by the other spouse: both husbands and wives
reported that their levels of work stress on one day were negatively associated with the
intimacy experienced by their spouse later that day. Combining the spillover and crossover
effects, it appears that spouses were not as intimate on days after either or both experienced
high work stress, but experienced more intimacy after low work stress days. These data
replicate the results of Bolger et al. (1989).
Stress Variables on One Dav Predicting Husbands’ Marital Interactions the Same Dav
Intimacy Physical
Conflict
Verbal
Conflict
Withdrawal
from Partner
Withdrawal
by Partner
Variable B
(SEE)
B
fSE B1
B
(SEB)
B
(SEB)
B
(SEB)
Husbands’ Report
Work stress -.13** .02 -.09* .01 .01
(•04) (.04) (.04) (.04) (.04)
Wives’ Report
Work stress -.13** -.04 .07 -.04 .06
(.04) (.04) (.04) (.04) (.04)
*£<.05. **£<.01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
Table 11. Work-to-home Stress Contagion: Summary o f Regression Analysis for
Work Stress Variables on One Day Predicting Wives’ Marital Interactions the Same Day
Variable
Intimac
y
Physical
Conflict
Verbal
Conflict
Withdrawal
from Partner
Withdrawal
by Partner
B
(SEB)
B
(SEB)
B
(SEB)
B
(SEB)
B
(SEB)
Husbands’ Report
Work stress -.16*** .01 -.00 -.02 -.00
(.04) (.04) (-04) (-04) (.04)
Wives’ Report
Work stress
_ 15***
-.06 -.02 -.00 .10*
(.04) (.04) (.04) (.04) (-04)
*g<.05. ***£<.001.
Second, the results shown in Table 11 indicate a significant crossover effect from
wives’ work stress to withdrawal by husbands. Wives reported a significant negative
association between their levels of work stress and the withdrawal they perceived in their
husbands later that day. That is, wives may have perceived higher withdrawal in their
husbands after their (the wives’) high work stress days, and lower withdrawal after low work
stress days. The dyadic adjustment effect (Bolger et al., 1989; Repetti, 1989) may explain
this finding. Husbands may have adjusted to wives’ high stress work days by withdrawing
from marital interactions, thus allowing the wives to recuperate. This type of adjustment is
different from one detailed by Repetti (1989), who found that wives adjusted to husbands’
high stress work days by providing more social support than usual.
The data also provide evidence contrary to a prediction made by Hypothesis One.
The results shown in Table 10 indicate a spillover effect from husband’s work stress to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
marital verbal conflict. This spillover effect, however, is in a direction opposite to the one
predicted: husbands reported a significant negative association between their levels of daily
work stress and the marital verbal conflict they experienced later that day. This suggests
husbands may have perceived less verbal conflict in the marital relationship on days when
they experienced high work stress. A possible explanation for this result is that husbands
may use withdrawal from the marital relationship as one way to cope with high work stress.
They may thus decrease verbal interactions with their wives, leading to less verbal conflict.
Neither spillover nor crossover measures of withdrawal, however, confirm this explanation,
and thus this result remains puzzling.
It appears, then, that spillover and crossover hypotheses regarding the effects of work
stress on the marital relationship were confirmed for the intimacy dimension of that
relationship. The data showed some evidence of dyadic adjustment in husbands, who
withdrew more on days when their wives experienced high work stress. Marital conflict was
not significantly associated with work stress. The present data, because they do not confirm
all the marital relationship predictions made by Hypotheses One and Three, stand in contrast
to previous theoretical discussions (e.g., Pleck, 1977) and empirical investigations (e.g.,
Repetti, 1989). Bolger et al. (1989), however, also found less consistent evidence for work-
to-home stress contagion than previous theory and research suggested.
Work-to-Home Stress Contagion: The Parent-Child Relationship
The data provide support for two predictions made by Hypothesis One. The results
shown in Tables 12 and 13 indicate a significant spillover effect from work stress to affection
and involvement in the parent-child relationship: both husbands and wives report that their
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
levels of work stress on one day were negatively associated with the affection and
involvement between themselves and the target child later that day. It appears, then, that
parents were less affectionate and involved with their children after high work stress days,
and more affectionate and involved after low work stress days.
Table 12. Work-to-home Stress Contagion: Summary of Regression Analysis for Work
Stress Variables on One Dav Predicting Hasbands’ Parent-child Relationship the Same Dav
Affection Conflict Involvement
Variable B SEB B SEB B SE B
Husbands’ Report
Work stress -.27*** .04 -.11** .04 -.32*** .04
Wives’ Report
Work stress -.03 .04 -.03 .04 -.03 .04
**E<.01. ***£<.001.
Table 13. Work-to-home Stress Contagion: Summary of Regression Analysis
for Work Stress Variables on One Dav Predicting Wives’ Parent-child Relationship
the Same Day
Affection Conflict Involvement
Variable B SEB a SEB B SEB
Husband’s Report
Work stress -.04 .04 .08 .04 .02 .04
Wife’s Report
Work stress
_ 24***
.04 .03 .04 -.27*** .04
* * * £ < . 001.
The data also provide evidence contrary to a prediction made by Hypothesis One.
The results shown in Table 12 indicate a spillover effect from husbands’ work stress to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
parent-child conflict. This spillover effect, however, is in a direction opposite to the one
predicted: husbands reported a significant negative association between their levels of daily
work stress and the parent-child verbal conflict they experienced later that day. This
suggests husbands experienced less conflict in their relationship with the target child on high
work stress days, but more conflict on low work stress days. A possible explanation for this
result relates to the previously discussed negative association between husbands’ work stress
and involvement in the parent-child relationship. On high work stress days, husbands had
fewer interactions with the target child, and thus the potential for conflict may have
decreased. In contrast, on low work stress days, husbands had more interactions with the
target child, and thus the potential for conflict may have increased.
It appears, then, that hypotheses regarding spillover effects from work stress to the
parent-child relationship were confirmed for the affection and involvement dimensions of
that relationship. Work stress was predictive of parent-child conflict in husbands but not in
wives. This aspect of the data provides some support for the theory that stress contagion is
more powerfiil among men than women (e.g., Pleck, 1977). The data did not support the
predictions made by Hypothesis Three: spouses reported no crossover effects from work
stress to the parent-child relationship.
Home-to-Work Stress Contagion: The Marital Relationship
The data do not support the predictions made by Hypothesis Two and Hypothesis
Four. The results shown in Tables 14 and 15 indicate no spillover or crossover effects from
marital interactions to work stress. It appears, then, that marital interactions on one day did
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
not have significant effects on work stress for either spouse the following day. This finding
is in direct contrast to previous theory and research in the field. For example, Pleck (1977)
hypothesized that home-to-work stress spillover effects would be stronger in husbands than
in wives; both Crouter (1984) and Bolger et al. (1989) found data to support this prediction.
Analvsis for Variables Within the Marital Relationship on One Dav
Predicting Husbands’ Work Stress Outcomes the Following Dav
Variable B SEB
Husbands’ Report
Intimacy -.02 .05
Physical conflict -.00 .04
Verbal conflict -.06 .05
Withdrawal from partner -.01 .05
Withdrawal by partner -.04 .04
Wives’ Report
Intimacy .05 .05
Physical conflict .05 .04
Verbal conflict -.02 .05
Withdrawal from partner .03 .05
Withdrawal by partner -.03 .05
Home-to-Work Stress Contagion: The Parent-Child Relationship
The data provide support for a prediction made by Hypothesis Four. The results
shown in Table 16 suggest a significant crossover effect from wives’ affection in the parent-
child relationship to husbands’ work stress the following day. Specifically, husbands
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
reported higher work stress on days after highly affectionate wife-child interactions, and
lower work stress on days after less affectionate wife-child interactions.
Analvsis for Variables Within the Marital Relationship on One Dav
Predicting Wives’ Work Stress Outcomes the Following Dav
Variable B SEB
Husbands’ Report
Intimacy .03 .05
Physical conflict .03 .04
Verbal conflict -.01 .05
Withdrawal from partner .01 .05
Withdrawal by partner .07 .03
Wives’ Report
Intimacy -.03 .05
Physical conflict -.03 .04
Verbal conflict -.07 .05
Withdrawal from partner .02 .05
Withdrawal by partner -.01 .05
The data also provide support for a prediction made by Hypothesis Two. The results
shown in Table 17 suggest a significant spillover effect from wives’ involvement in the
parent-child relationship to their work stress the following day. Specifically, wives reported
higher daily work stress following a day of many interactions with the target child, and lower
daily work stress following a day of fewer interactions.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
Table 16. Home-to-work Stress Contagion: Summary o f Regression
Analysis for Variables Within the Parent-child Relationship on One
Dav Predicting Husbands’ Work Stress Outcomes the Following Dav
Variable B SEB
Husbands’ Report
Affection with child
Conflict with child
Involvement with child
Wives’ Report
Affection with child
Conflict with child
Involvement with child
*p < .05.
Table 17. Home-to-work Stress Contagion: Summary of Regression
Analysis for Variables Within the Parent-child Relationship on One
Day Predicting Wives’ Work Stress Outcomes the Following Day
Variable B SE B
Husband’s Report
Affection with child -.09 .05
Conflict with child .03 .04
Involvement with child .10* .05
Wife’s Report
Affection with child .08 .04
Conflict with child .07 .04
Involvement with child -.13** .04
*P<.05. **p<.01.
.05
.00
.02
.04
.05
.05
-.08*
.02
.05
.04
.04
.04
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
Finally, the data provide evidence contrary to a prediction made by Hypothesis Four
and to theoretical discussions elsewhere (e.g., Swick, 1988). The results shown in Table 16
indicate a crossover effect from husbands’ involvement in the parent-child relationship to
wives’ work stress. This crossover effect, however, is in a direction opposite to the one
predicted: spouses reported a significant negative association between wives’ levels of daily
work stress and husband-child involvement later that day. This suggests wives experienced
high work stress on days following many interactions between husbands and target children,
and low work stress on days following few interactions. One explanation for this effect may
be that dual-earner parents can spend limited time with their children, and so husbands and
wives in this sample may not have been involved with their children simultaneously. Thus,
when husbands were highly involved with the couples’ children, wives may not have spent
time with them. This, in turn, may have led wives to perceive the familial interaction as
negative, and, in line with spillover hypotheses, to experience high work stress the following
day. This explanation bears some scrutiny given that wives reported that high involvement
mother-child evenings were associated with lower work stress the following day.
It appears, then, that hypotheses regarding spillover and crossover effects from the
parent-child relationship to work stress were, in large part, not confirmed. As with the
marital relationship, the data show that parent-child interactions on one day did not have
significant effects on work stress for either spouse the following day. The significant effects
that were found suggest that stress contagion from the parent-child relationship, if it does
occur, is more likely to do so in wives than in husbands. Again, this finding is not
concordant with previous theory and research (e.g., Bolger et al., 1989; Crouter, 1984; Pleck,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
1977), though it should be noted that no previous study considered the effects of the parent-
child relationship on work stress.
Discussion
The major purpose of this study was to test the prediction that work stress and family
relationships affect each other on a daily basis. In pursuing this purpose, the paper
attempted, first, to replicate the findings of previous empirical studies (e.g., Bolger et al.,
1989; Crouter et al., 1989; Doumas, 1993), and, second, to extend theoretical and empirical
models of stress contagion to the parent-child relationship.
The study’s work-to-home spillover hypotheses were partially confirmed. Daily
work stress and one dimension of the marital relationship (intimacy) were negatively
associated in both husbands and wives. Daily work stress and two dimensions of the parent-
child relationship (affection and involvement) were negatively associated in both husbands
and wives. These results strongly suggest a spillover effect between work and home. In this
respect, the present study replicates Bolger et al.’s (1989) findings; it does not, however,
replicate earlier findings that suggested high work stress leads to greater conflict and
withdrawal between spouses.
The study’s home-to-work spillover hypotheses were not confirmed. Only one
dimension of family relationships seemed to have an impact on subsequent work stress:
wives’ daily work stress was negatively associated with their involvement in the parent-child
relationship on the previous day. These results do not replicate Bolger et al.’s (1989)
findings, which suggested that home-to-work stress contagion, especially in husbands, is a
powerful effect.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
The study’s work-to-home and home-to-work crossover hypotheses were not
confirmed. These results do not replicate the findings of either Bolger et al. (1989) or
Crouter et al. (1989).
In summary, this study confirms the hypothesis that work stress spills over to the
home. The results do not, however, support the notion that home-to-work stress contagion
is stronger than work-to-home stress contagion (see Bolger et al., 1989), and provide scant
evidence for crossover effects (see Crouter et al., 1989). It appears, then, that the present
study fails to provide support for previous theoretical work and to replicate previous research
findings.
This may be at least partially due to several limitations in this study’s research
design. Perhaps the most important of these is that data were collected as part of a larger
study and not specifically for the purposes of this study. Thus the variables available were
limited, and their definition was not ideal for the current purposes. For instance, few
questionnaire items could contribute directly to the construction of a work stress index; the
data analysis used all suitable items, but still could not assess the effects of time pressure and
deadlines on daily work stress. Similarly, some respondents (e.g., teachers who grade papers
at night, physicians who attend weekend workshops) might have had trouble accurately
answering questions about workload and paid hours worked. Finally, the data collection
procedures made it impossible to know whether participants were responding accurately and
promptly, and what effects this may have had on the data analysis.
Another explanation for the failure to replicate earlier findings may center on the
inclusion of the parent-child relationship in the present study. Previous studies did not
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
include this relationship, or the presence or absence of children, as variables. The present
results may suggest that the childless couples experience different stress contagion processes
than couples with children. Thus, the most important contribution of this study to the
literature may be its introduction of the parent-child relationship into the study of the work-
family interface. Future research might explore whether and how the parent-child
relationship moderates the association between daily work stress and the marital interaction.
Future studies in this field should employ robust longitudinal designs, and larger
sample sizes and longer periods of data collection than was the case in the present study. In
addition, investigators might improve on the current study by having as their sole and precise
focus the questions this study attempted to answer. More sensitive, better defined measures
of work stress and the dimensions of the marital and parent-child relationships will aid
attempts to disconfirm theoretically principled predictions.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Notes
1 Throughout this article the term “work” is used to refer to paid work in the labor
market, where the worker performs an occupational role in contrast to the family roles he
performs at home. Thus, the term “dual-eamer families,” as used here, refers to those
husband-wife families in which both partners regularly participate in paid market work.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
51
Barling, J. (1984). Effects of husbands’ work experiences on wives’ marital
satisfaction. Journal of Social Psychology. 124.219-225.
Barling, J., & Rosenbaum, A. (1986). Work stressors and wife abuse. Journal of
Applied Psychology. 71. 346-348.
Baruch, G.K., Biener, L., & Barnett, R.C. (1987). Women and gender in research on
work and family stress. American Psychologist. 42.130-136.
Bates, F.L. (1956, December). A conceptual analysis of group structure. Social
Forces. 103-111.
Blood, R., & Wolfe, D. (1960). Husbands and wives. New York: Macmillan.
Bolger, N., DeLongis, A., Kessler, R.C., & Wethington, E. (1989). The contagion
of stress across multiple roles. Journal of Marriage and the Family. 51. 175-183.
Brody, G.H., Pellegrini, A.D., & Sigel, I.E. (1986). Marital quality and mother-child
and father-child interactions with school-aged children. Developmental Psychology. 22.
291-296.
Bronfenbrenner, U., & Crouter, A.C. (1982). Work and family through time and
space. In S.B. Kammerman & C.D. Hayes (Eds.), Families that work: Children in a changing
world. Washington, DC: National Academy Press.
Brown, R.W. (1986). Social psychology: The second edition. New York: The Free
Press.
Burke, R.J. (1982). Occupational demands on administrators and spouses’
satisfaction and well-being. Psychological Reports. 51. 823-836.
Burke, R.J., & Weir, T. (1976). Relationship of wives’ employment status to
husband, wife and pair satisfaction and performance. Journal of Marriage and the Family.
38,279-287.
Burman, B., John, R.S., & Margolin, G. (1992). Observed patterns of conflict in
violent, nonviolent, and nondistressed couples. Behavioral Assessment. 14.15-37.
Caspi, A., Bolger, N., & Eckenrode, J. (1987). Linking person and context in the
daily stress process. Journal of Personality and Social Psychology. 5 2 .184-195.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
Chan, C.G.L., & Margolin, G. (in press). The relationship between dual-earner
couples’ daily work mood and their daily marital interaction.
Clark, R.A., Nye, F.I., & Gecas, V. (1978). Husbands’ work involvement and marital
role performance. Journal of Marriage and the Family. 40. 9-21.
Crouter, A.C. (1984). Spillover from family to work: The neglected side of the work-
family interface. Human Relations. 37.425-442.
Crouter, A.C., Perry-Jenkins, M., Huston, T.L., & Crawford, D.W. (1989). The
influence of work-induced psychological states on behavior at home. Basic and Applied
Social Psychology. 10.273-292.
Doumas, D. (1993). The relationship between daily marital interaction and daily
stress and daily health behaviors: An extension of the work-familv spillover model.
Unpublished doctoral dissertation, University of Southern California, Los Angeles.
Easterbrooks, M.A., & Emde, R.N. (1988). Marital and parent-child relationships:
The role of affect in the family system. In R.A. Hinde & J. Stevenson-Hinde (Eds.),
Relationships within families: Mutual influences. Oxford: Clarendon.
Eckenrode, J. (1984). Impact of chronic and acute stressors on daily reports of mood.
Journal of Personality and Social Psychology. 46. 907-918.
Gottman, J.M. (1981). Time series analysis: A comprehensive introduction for social
scientists. Cambridge, England: Cambridge University Press.
Gregson, R.A.M. (1983). Time series in psychology. Hillsdale, NJ: Erlbaum.
Hedges, S.M., Jandorf, L., & Stone, A.A. (1985). Meaning of daily mood
assessments. Journal of Personality and Social Psychology. 48.428-434.
Hoffman, L.W., & Hoffman, H.J. (1985). The lives and adventures of dual-career
couples. Family Therapy. 12.123-149.
Judge, G.G., Griffiths, W.E., Hill, R.C., Lutkepohl, H., & Lee, T.C. (1985). The
theory and practice of econometrics. New York: Wiley.
Kanter, R.M. (1977). Work and family in the United States: A critical review and
agenda for research and policy. New York: Sage.
Kessler, R.C., & McRae, J.A. (1982). The effect of wives’ employment on the
mental health of married men and women. American Sociological Review. 47. 216-227.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
Kingston, P.W., & Nock, S.L. (1985). Consequences of the family work day.
Journal of Marriage and the Family. 47.619-629.
Ladewig, B.H., & White, P.N. (1984). Dual-earner marriages: The family social
environment and dyadic adjustment. Journal of Family Issues. 5 .343-362.
Locksley, A. (1980). On the effects of wives’ employment on marital adjustment and
companionship. Journal of Marriage and the Family. 42. 337-346.
Loscocco, K.A., & Roschelle, A.R. (1991). Influences on the quality of work and
nonwork life: Two decades in review. Journal of Vocational Behavior. 3 9 .182-225.
Malmaud, RK. (1980). Work and marriage: The two-profession couple. Ann Arbor,
MI: UMI Research Press.
Margolin, G. (1989). Adult Home Data Questionnaire. Unpublished measure,
University of Southern California, Los Angeles.
Margolin, G., & John, R.S. (1988). Marital conflict: Interpersonal and intrapersonal
factors. NIMH Grant No. MH36595.
Margolin, G., John, R.S., & Burman, B. (1993). Family factors and children’s
maladjustment: Daily variability. Paper presented at the 60th Annual Meeting of the Society
for Research in Child Development.
Margolin, G., John, R.S., & O’Brien, M. (1989). Sequential affective patterns as a
function of marital conflict style. Journal of Social and Clinical Psychology. 8 .45-61.
Marks, S.R. (1977). Multiple roles and role strain: Some notes on human energy,
time, and commitment. American Sociological Review. 42.921-936.
McKinley, D.G. (1964). Social class and family life. Glencoe, NY: Family Press.
Nesselroade, J.R. (1988). Some implications of the trait-state distinction for the
study of development over the life span: The case of personality. In P. Baltes, D.
Featherman, & R. Lemer(Eds.), Life-span development and behavior fVol. 8, pp. 163-189).
Hillsdale, NJ: Erlbaum.
Neter, J., Wasserman, W., & Kutner, M.H. (1983). Applied linear regression models.
Homewood, IL: Richard D. Irwin.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
Ostrom, C.W. (1990). Time series analysis: Regression techniques (2nd ed.). Sage
University Paper series on Quantitative Applications in the Social Sciences, 07-009.
Newbury Park, CA: Sage.
Piotrkowski, C.S. (1979). Work and the family system. New York: The Free Press.
Piotrkowski, C.S., & Crits-Cristoph, P. (1982). Women’s jobs and family
adjustment. In J. Aldous (Ed.), Two paychecks: Life in dual-earner families (pp. 105-127).
Beverly Hills, CA: Sage.
Piotrkowski, C.S., Rapoport, R.N., & Rapoport, R. (1987). Families and work. In
M.B. Sussman & S.K. Steinmetz (Eds.), Handbook of marriage and the family. New York:
Plenum Press.
Piotrkowski, C.S., & Repetti, R.L. (1984). Dual-eamer families. Marriage and
Family Review. 7. 99-124.
Pleck, J.H. (1977). The work-family role system. Social Problems. 24.417-427.
Pleck, J.H. (1985). Working wives, working husbands. Beverly Hills, CA: Sage.
Pleck, J.H., & Staines, G.L. (1985). Work schedules and family life in two-earner
couples. Journal of Family Issues. 6. 61-82.
Rapoport, R., & Rapoport, R. (1965). Work and family in contemporary society.
American Sociological Review. 30. 381-394.
Renne, K.S. (1970). Correlates of dissatisfaction in marriage. Journal of Marriage
and the Family. 32. 54-67.
Repetti, R.L. (1987). Linkages between work and family roles. In S. Oskamp (Ed.),
Applied social psychology annual: Family process and problems (pp. 98-127). Beverly
Hills, CA: Sage.
Repetti, R.L. (1989). Effects of daily workload on subsequent behavior during
marital interaction: The roles of social withdrawal and spouse support. Journal of
Personality and Social Psychology. 57.651-659.
Ridley, C. A. (1973). Exploring the impact of work satisfaction and involvement on
marital interaction when both partners are employed. Journal of Marriage and the Family.
35* 229-237.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
Staines, G.L. (1980). Spillover versus compensation: A review on the literature on
the relationship between work and nonwork. Human Relations. 33. 111-129.
Swick, K.R. (1988). Parental efficacy and involvement: Influences on children.
Childhood Education. 65.37-42.
Thoits, P.H. (1983). Multiple identities and psychological well-being. American
Sociological Review. 4 8 .174-187.
Yogev, S. (1986). Relationships between stress and marital satisfaction among dual-
earner couples. Women and Therapy. 5. 313-330.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
Appendix A
Adult Home Data Questionnaire
Family ID#____________Day of Interaction___________ Date Filled Out___________
DAILY STRESS: (“Work refers to paid and unpaid work)
1. My stress level at work today was: a) high b) moderate c) low d) none
2. My work load today was: a) high b) average c) low d) none
3. My enjoyment of work today was: a) high b) average c) low d) none e) did no work
4 .1 was irritated/annoyed by someone at work today: a) a lot b) a little c) not at all
5. Someone at work was irritated/annoyed with me: a) a lot b) a little c) not at all
6 .1 still felt stress or pressure from work when I was home with my family:
a) a lot b) some c) a little d) none
7. My productivity (at work and home) today was: a) good b) moderate c) poor
8. My relations with people outside the home (e.g., friends, doing errands) were:
a) good b) moderate c) poor d) none
9. My contact with people I like (other than my family) was:
a) high b) moderate c) low d) none
10. Today I felt: a) energetic b) neutral c) listless
11. Today the stress I felt from my family was: a) high b) moderate c) low d) none
12. Today the stress I felt from other than my family (work, friends, activities) was:
a) high b) moderate c) low d) none
13. Overall my day outside my family was:
a) very positive b) positive c) neutral d) negative e) very negative
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
14. Today I felt stressed by financial concerns: a) a lot b) some c) a little d) none
15. Today I felt stressed by decisions: a) a lot b)some c) a little d)none
16. Today I felt stressed by health problems (my own or my family members):
a) a lot b) some c) a little d) none
17. Today I felt stressed by issues of safety for myself and my family:
a) a lot b) some c) a little d) none
18. Today I felt stressed by tasks needing my attention: a) a lot b)some c) a little d) none
19. Today I felt stressed because some item (appliance, car, etc.) was not working or I lost
something: a) a lot b) some c) a little d) none
CHILD’S BEHAVIOR:
20. The following describe my child today. Circle all that apply.
a) happy k) screamed u) fought with a friend
b) sad I) helpful v) fought with a sibling
c) scared m) had a good day at school w) did homework
d) angry n) threw temper tantrum x) destroyed an object
e) anxious o) jealous of a sibling belonging to another
f) calm p) disobeyed parent y) cruel
g) cranky or whiny q) demanded attention z) lied
h) argumentative r) played well alone aa) swore
i) cooperated s) played well with friends bb) stole
j) friendly t) sassy to parent cc) moody
dd) cried
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
PARENT-CHILD INTERACTION
21. How much contact did you have with your child today?
a) none b) < 1 hour c) 1-3 hours d) 4-6 hours e) > 6 hours
22. How much love or affection did you show your child today?
a) none b) a little c) some d) quite a bit e) a lot
23. How angry were you at your child today?
a) none b) a little c) some d) quite a bit e) a lot
24. How angry was your child at you today?
a) none b) a little c) some d) quite a bit e) a lot
25. How much did you punish your child today?
a) none b) a little c) some d) quite a bit e) a lot
26. Were you too busy to do something your child wanted you to do today?
a) none b) a little c) some d) quite a bit e) a lot
27. How much did you nag your child today?
a) none b) a little c) some d) quite a bit e) a lot
28. How irritated were you with your child today?
a) none b) a little c) some d) quite a bit e) a lot
29. How often did you warn your child she or he might be punished?
a) none b) a little c) some d) quite a bit e) a lot
30. How much did you yell at your child today?
a) none b) a little c) some d) quite a bit e) a lot
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
31. How much did you tell your child to stop doing something?
a) none b) a little c) some d) quite a bit e) a lot
32. How much did you hug or kiss your child today?
a) none b) a little c) some d) quite a bit e) a lot
33. How much did you ask your child to do something (chore) more than once?
a) none b) a little c) some d) quite a bit e) a lot
34. Overall my day with my child was:
a) very positive b) positive c) neutral d) negative e) very negative
MARITAL INTERACTION
35. How much contact did you have with your spouse today?
a) none b) < 1 hour c) 1-3 hours d) 4-6 hours e) > 6 hours
36. Today I expressed anger or irritation at my partner:
a) none b) a little c) some d) quite a bit e) a lot
37. Today my partner expressed anger or irritation at me:
a) none b) a little c) some d) quite a bit e) a lot
38.1 felt distant or withdrawn from my partner:
a) not at all b) a little c) some d) quite a bit e) a lot
39. My partner seemed distant or withdrawn from me:
a) not at all b) a little c) some d) quite a bit e) a lot
40. Today I hit, pushed, or shoved my partner:
a) not at all b) a little c) some d) quite a bit e) a lot
41. Today my partner hit, pushed, or shoved me:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
a) not at all b) a little c) some d) quite a bit e) a lot
42. My partner and I kissed and hugged each other today:
a) not at all b) a little c) some d) quite a bit e) a lot
43. My partner and I had a good conversation today:
a) not at all b) a little c) some d) quite a bit e) a lot
44. Today I nagged my partner: a) not at all b) a little c) some d) quite a bit e) a lot
45. Today my partner nagged me: a) not at all b) a little c) some d) quite a bit e) a lot
46. My partner ignored my wishes or needs:
a) not at all b) a little c) some d) quite a bit e) a lot
47.1 ignored my partner’s wishes or needs:
a) not at all b) a little c) some d) quite a bit e) a lot
48.1 took my partner’s feelings lightly:
a) not at all b) a little c) some d) quite a bit e) a lot
49. My partner took my feelings lightly:
a) not at all b) a little c) some d) quite a bit e) a lot
50. My partner and I disagreed about a child-related issue:
a) not at all b) a little c) some d) quite a bit e) a lot
51. My partner and I disagreed about an issue unrelated to children:
a) not at all b) a little c) some d) quite a bit e) a lot
52. Overall my day with my partner was:
a) very positive b) positive c) neutral d) negative e) very negative
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
61
53) Did you and your spouse have any conflictual interaction today in the presence of your
child?____________Yes__________ No
54. If you answered yes to question 53, please indicate your child’s reaction to the
interaction. Circle all that apply.
a) Cried g) Took sides
b) Showed no reaction h) Misbehaved or had a tantrum
c) Became angry i) Tried to make peace
d) Seemed to feel distressed, sad, or frightened j) Became unusually
e) Left the room well-behaved
f) Listened or watched k) Picked a fight
55. Did you or your spouse show affection towards one another today in the presence of your
child?____________ Yes____________No
56. If you answered yes to question 53, please indicate your child’s reaction to the
interaction. Circle all that apply.
a) Seemed embarrassed d) Watched
b) Displayed jealousy e) Seemed comfortable and happy
c) Showed no reaction f) Joined in to also get some affection
57. How much time did you spend in your car today?
a) none b) < 1 hour c) 1-2 hours d) 2-4 hours e) > 4 hours
58. How many hours did you sleep last night?
a) <2 hours b) 2-5 hours c) 5-7 hours d) 7-9 hours e) > 9 hours
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
62
59. Did you have trouble sleeping?
a) not at all b) a little c) some d) a lot
60. How many alcoholic drinks did you consume today?
a) none b)one c)two d) three e)>3
61. How many cigarettes did you smoke today?
a) none b) 1-10 c) 11-20 d)>20
62. How many meals did you eat today?
a) none b) one c) two d) three e) > 3
63. Did you overeat today?
a) a lot b) some c) a little d) not at all
64. Did you experience physical discomfort today?
a) none b) slight c) moderate d) a lot
65. Did you take any non-prescription medication today? a) yes b) no
66. Did you take any prescribed pain medicines, sedatives, or stimulants? a) yes b) no
67. Did you take any recreational drugs today? a) yes b) no
68. How much time did you spend exercising today?
a) none b) < 30 minutes c) 30-60 minutes d) >1 hour
69. How many paid hours did you work today?
a) none b) < 4 hours c) 4-8 hours d) 8-12 hours e) > 12 hours
70. How many unpaid hours did you work today?
a) none b) < 4 hours c) 4-8 hours d) 8-12 hours e) > 12 hours
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71. How many hours did you spend watching TV or VCR today?
a) none b) < 1 hour c) 1-3 hours d) 3-6 hours e) > 6 hours
72. How much time did you read for pleasure today?
a) none b) < 1 hour c) 1-3 hours d) > 3 hours
73. Did you have sexual intercourse within the last 24 hours? a) yes b) no
74. How much time did you spend relaxing today?
a) none b) < 1 hour c) 1-3 hours d) 3-6 hours e) > 6 hours
75. How much time did your child spend watching TV or VCR today?
a) none b) < 1 hour c) 1-3 hours d) 3-6 hours e) > 6 hours
PARENT-CHILD DAILY LEISURE ACTIVITIES - Did you and your child do any of the
following activities TOGETHER today? Circle NO or YES. If YES, rate how pleasurable
you think the activity was FOR YOU.
Pleasurable Rating
Not PI. = » Very PI.
76) Playing NO Y ES- 1 2 3 4
77) Shopping NO Y ES- 1 2 3 4
78) Doing homework NO Y ES- 1 2 3 4
79) Reading NO Y ES- 1 2 3 4
80) Taking a walk NO Y ES- 1 2 3 4
81) Dance, music, art NO Y ES- 1 2 3 4
82) Discussing child’s day NO Y ES- 1 2 3 4
83) Cooking, chores NO Y ES- 1 2 3 4
84) Sports, board games NO Y ES- 1 2 J 4
85) Going out to an event NO Y ES- 1 2 3 4
86) Eating together NO Y ES- 1 2 3 4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
Pleasurable Rating
Not PL - Very PI.
87) Watching TV or VCR NO YES- 1 2 3 4
88. If you know, please list the shows your child watched on TV or VCR today:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
Appendix B
Common Errors Sheet
1) Not filling out “Day of Activity” and “Day Reported” correctly.
• Date of Activity = the day on which the events you are describing have occurred.
• Date Reported = the day you are filling out the questionnaire.
2) Skipping question 54 and question 56.
• A “yes” response on question 53 requires a response to question 54. Similarly, a
“yes” response to question 55 requires a response to question 56.
3) Forgetting to circle “no” responses on questions 76-87.
• Each one of these questions must be answered. Either a “yes” or a “no” must be
circled. A “yes” response further requires that you rate how pleasurable the activity
was.
4) Skipping questions: It is suggested that you go over the questionnaire before sealing it and
complete any items that have been skipped.
5) Circling more than one response: Only one response per question is acceptable.
6) Forgetting rules regarding days of no contact with spouse or child.
• Contact = any interaction at all. Thus, even if you do not see your spouse or child,
consider whether, for example, you spoke on the phone or exchanged notes.
• On days of no contact at all, the response to questions 34 and 52 should be “c.”
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
66
7) Leaving question 88 blank.
• If you do not know or cannot remember the name(s) of the television program(s)
that your child watched, write “I don’t know” in response to this question, rather than
leaving it blank, so that we know that you did not skip the question.
Bonus
The postmark on the envelope must be the next day after the day of activity. So if
you are reporting on the 15th, the envelope must be postmarked the 16th. If all forms are
received on time, a $5 bonus will be added to the $20 earned for the procedures.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The associations among family stress, friendship quality, and romantic quality

PDF

Affection and conflict in family relationships

PDF

Parents' psychological distress and children's temperament as moderators of the effects of daily work stress on family interactions

PDF

Parental frequency and satisfaction with children's daily activities

PDF

The effects of community violence on the well-being of Latino mothers and their preschool children

PDF

Television viewing and creativity in young children

PDF

Stress and coping among gay older adults

PDF

The relationship between changes in the type and frequency of contact in open and closed adoptions and behavioral outcomes of adopted children

PDF

Genetic cross-relationships between criminal behavior and severe alcohol-related problems

PDF

Emotion regulation as a mediator between family-of-origin aggression and marital aggression

PDF

Content analysis of articulated thoughts of chronic worriers

PDF

Values and their relationships to HIV-related behavior, attitudes, and social norms

PDF

The relationship of hardiness and racial identity attitudes to the articulated thoughts and psychological impact of simulated racism encounters

PDF

Depression and suicidality in Latino adolescents: A study of acculturation and gender role beliefs

PDF

Trajectories of physical and emotional partner aggression in marriage

PDF

Quality of life and positive expectations in women with recurrent breast cancer

PDF

Social-cognitive and behavioral attributes of subtypes of aggressors and victims in children's peer groups

PDF

The relationship between acculturation and the household structure of older Mexican Americans in Los Angeles

PDF

Violence exposure, friendship, and social adjustment among Head Start preschoolers

PDF

Triple threat: Community violence as a predictor for post traumatic stress disorder and the associated features of drug use and depressive symptoms among Latino youth

Asset Metadata

Creator Thomas, Kevin Garth Flusk (author)

Core Title The relationship between work stress and family interaction in dual-earner couples

Degree Master of Arts

Degree Program Psychology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,Psychology, clinical,psychology, social,sociology, individual and family studies

Language English

Contributor Digitized by ProQuest (provenance)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-6652

Unique identifier UC11341102

Identifier 1381610.pdf (filename),usctheses-c16-6652 (legacy record id)

Legacy Identifier 1381610.pdf

Dmrecord 6652

Document Type Thesis

Rights Thomas, Kevin Garth Flusk

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA