Predictors of Parenting Stress for Abusive
and Nonabusive Mothers
Andrea V. McPherson Æ Æ Kristen M. Lewis Æ Æ
Amy E. Lynn Æ Æ Mary E. Haskett Æ Æ Tara S. Behrend
? Springer Science+Business Media, LLC 2008
and nonabusive mothers (n = 86) using linear regression analyses. Predictors in the
model included (a) the degree to which mothers were bothered by child misbe-
havior, (b) mothers’ general psychological functioning, and (c) observed child
behavior during parent–child interactions. Whether abuse status moderated the
relations between each predictor and parenting stress was also explored. Results
indicated that mothers’ psychological functioning significantly predicted parenting
stress; however, neither mothers’ intolerance for their children’s misbehavior nor
observed child behavior were significant predictors of parenting stress in the
regression model. A test of moderation revealed a significant interaction between
parental intolerance and abuse status such that intolerance predicted parenting stress
level only for abusive mothers. A comparison of correlations indicated that abusive
mothers’ level of parenting stress was more closely related to their intolerance for
child conduct problems than to the child’s behavior during play with their mothers.
Findings are discussed in terms of implications for interventions to reduce parenting
stress experienced by nonabusive and abusive mothers.
We examined a model of parenting stress for abusive mothers (n = 80)
Psychological distress ? Child adjustment
Parenting stress ? Abusive mothers ? Parental tolerance ?
Stress that arises from the demands of parenthood is linked to several aspects of
parenting, including low parental warmth (Rogers 1993), negative and controlling
behavior (Bigras et al. 1996), and harsh discipline practices (Pinderhughes et al.
A. V. McPherson ? K. M. Lewis ? A. E. Lynn ? M. E. Haskett (&) ? T. S. Behrend
Department of Psychology, North Carolina State University, Box 7650, Raleigh, NC 27695, USA
J Child Fam Stud
2000). Because stress has a negative impact on parenting quality, there has been an
effort to achieve an understanding of determinants of parenting stress. To gain such
knowledge, investigators have examined sources of parenting stress for clinical
populations such as adolescent parents (Budd et al. 2006) and parents of children
with ADHD (Fischer 1990). Other investigators have explored correlates of
normative parenting stress (Deater-Deckard and Scarr 1996). Few investigations,
however, have been designed to examine models of stress for samples that include
both high risk parents and those at relatively low risk for elevated levels of stress.
Because different pathways might lead to stress among parents at risk for strain in
the parenting role and parents with relatively few risk factors, this could be an
important research direction to inform intervention practices.
There is some suggestion in the literature that determinants of parenting stress
indeed might vary across groups of parents. For example, Creasey and Reese (1996)
suggested that correlates of stress might differ for parents of challenging children
and parents of children who are relatively easy to manage. Mash and Johnston
(1990) proposed that stress for parents of hyperactive children was based primarily
on child characteristics while stress for abusive parents arose from parental
characteristics such as perceptions of children and attributions for children’s
misbehavior. It also seems likely that determinants of stress might be different for
low-risk samples of generally warm, sensitive parents of children without behavior
problems and those at the extreme end of harsh discipline, such as abusive parents,
who tend to have children whose behavior can be challenging.
Although there are some exceptions (Holden and Banez 1996) many studies
indicate that abusive parents report elevated levels of parenting stress (e.g., Ethier
et al. 1995; Haskett et al. 2006). Furthermore, several studies indicate that high
parenting stress is related to increased child abuse potential (e.g., Holden and Banez
1996; Rodriguez and Green 1997). It follows logically from this literature that stress
would have a central role in many etiological models of abusive parenting. In fact,
cognitive-behavioral models (e.g., Azar 1986; Schellenbach et al. 1991) and social
information processing models (e.g., Milner 2003) place emphasis on stress as a
contributing factor to harsh, abusive parenting, and Wolfe’s (1999) transitional
model posits that parental low tolerance for stress initiates a chain of events toward
abusive parenting. Given the central role of stress in the etiology of child abuse,
attempting to identify unique predictors of stress for abusive parents is a worthy
The predictors of parenting stress examined in this study were chosen on the
basis of several existing models of parenting stress (see Deater-Deckard 2004).
Specifically, the proposed model includes components of parenting stress in
Abidin’s validated model (1992, 1995). Those components include children’s
behavioral adjustment and parental functioning (i.e., psychological well-being and
intolerance for children’s conduct problems). The inclusion of each predictor is
Disruptive child behavior problems are closely associated with high levels of
parenting stress (Ross et al. 1998). Abundant studies indicate that physically abused
children are characterized by externalizing problems such as aggression (e.g.,
Bolger and Patterson 2001), and abusive parents report that their children are
J Child Fam Stud
difficult to manage (Bradley and Peters 1991). Thus it is not surprising that abusive
parents tend to have elevated levels of parenting stress. There likely is a
bi-directional process that takes place between child behavior and parenting stress,
but for purposes of establishing a model for this investigation, we proposed that
children’s behavior during interactions with their mothers would predict mothers’
parenting stress. This proposed direction of influence is consistent with established
models of parenting stress (e.g., Abidin 1995) and with models set forth in prior
investigations (e.g., Williford et al. 2007).
There has been discussion in the literature about whether actual child behavior or
parent’s perceptions of children’s challenging behaviors are more closely tied to
levels of parenting stress. Studies designed to address this issue have produced
mixed findings. Creasey and Reese (1996) concluded that actual child behavior
(based on teachers’ reports) was more closely associated with parenting stress than
was low parental tolerance of deviant child behavior. Others (Bigras et al. 1996;
Creasey and Jarvis 1994) found parenting stress was related to parent reports of
child adjustment but was unrelated to child behavior during parent–child
interactions. Perhaps relations among stress, child behavior, and parent perceptions
vary depending on severity of stress and child behavior problems. Mash and
Johnston (1990) proposed that child behavior might be less relevant to stress for
abusive parents than their perceptions and tolerance of their children’s misbehavior;
actual child behavior might be a stronger predictor of stress for nonabusive than for
abusive parents. The authors based that proposition on the fact that abusive parents
tend to view their children as having more serious behavior and emotional problems
than do nonabusive parents even though observers’ perceptions of abused and
nonabused children often are similar (e.g., Mash et al. 1983).
According to Deater-Deckard (1998), stress related to the parenting role and
parents’ general psychosocial functioning are closely linked. Numerous studies
support that conclusion (e.g., Ethier et al. 1995; Gotlib and Hanunen 1992). For
example, Quittner et al. (1990) examined links among social support, parenting
stress, life event stress, and general psychological distress among mothers of deaf
children and matched controls. Stress related to the parenting role, not general life
stress, was most highly correlated with psychological distress. Although psycho-
logical distress and parenting stress are closely associated, some studies indicate
that they are in fact distinct indicators of functioning (e.g., Budd et al. 2006).
Webster Stratton (1990) claimed that positive mental health functioning could
protect against chronic stress so that parents with good mental health functioning
would be less impacted by situational stressors such as child disruptive behavior.
This investigation was designed to explore several determinants of parenting
stress. Our sample included abusive mothers with relatively high levels of parenting
stress so that we could examine the degree to which links between each predictor
and parenting stress varied across clinical and nonclinical samples. The model
examined in this study posited that children’s observed behavior during play with
their mothers, mothers’ intolerance for their children’s conduct problems, and
mothers’ general psychological functioning would predict level of parenting stress.
The link between each predictor and parenting stress was tested for moderation by
abuse status. We also examined the relative strength of links between parenting
J Child Fam Stud
stress of abusive mothers and two predictors—mothers’ intolerance and observed
child behavior. It was expected that the correlation between parenting stress and
intolerance would be significantly greater than the correlation between parenting
stress and child behavior.
Participants were 166 mothers and their children (51.8% female) between the ages
of 5–10 years (M = 7.2 years; SD = 1.5). Participants were drawn from a larger
study of the social adjustment of young children and were included in the current
study if the parent had completed the measure of parenting stress. There were no
differences in demographic characteristics between parents from the larger study
and those who participated in this investigation (all p values[.20). The larger
study included very few fathers, thus only mothers were involved in this study. Of
the 166 mothers, 80 had documented histories of physical abuse of the child in this
sample and 86 had no known history of abuse. The majority (70%) were African
American, 27.6% were Caucasian, and 2.4% were Hispanic or biracial. Nearly
37% of mothers were married or living with a partner. Although 20% had not
completed high school, 26% had a college degree. The full five-category range of
socioeconomic status based on the A. B. Hollingshead (Unpublished manuscript)
classification system was represented, with 33% at the two highest levels and 39%
at the two lowest levels. Mean parent age was 33.4 years (SD = 7.2). Abused
children (M = 7.4 years) were significantly older than comparison children
(M = 7.0 years), but that difference was probably not meaningful, and age was
unrelated to any study variables (all p values[.20). There were no group
differences on any other socio-demographic variables. Demographic characteristics
are provided in Table 1 to aid interpretation of findings.
To recruit abusive mothers, a description of the study was mailed to parents whose
children were eligible according to a review of the child protection registry. The
review was completed by staff of the child protection agency and the fourth author,
after she signed a confidentiality agreement with the agency. Comparison mothers
and children were recruited through flyers placed in neighborhoods where the abuse
participants resided. Interested parents called the project office and a psychosocial
interview, which included the Conflict Tactics Scale (CTS; Straus 1990), was
administered. The CTS is one of the most frequently used measures of family
violence and has very strong psychometric properties (Straus and Hamby 1997). To
participate, children had to be living with their parents at the time of data collection
and there could be no history of sexual abuse of the children or untreated maternal
substance abuse (based on parent report). Comparison parents were screened for
abuse history through the psychosocial interview, CTS (10 items related to
J Child Fam Stud
disciplinary tactics that could be considered abusive, including hitting, slapping,
punching, and shaking), and review of the child protection registry. Potential
comparison families were not enrolled in the study if there was any indication that
their child might have been physically abused based on any of these sources. If
families met criteria, data collection was scheduled. Families were assessed
individually in a university family clinic setting following informed consent.
Transportation and child care was provided to reduce those common barriers to
participation. Each parent who participated received $75. All methods and
procedures were approved by the university Institutional Review Board.
Table 1 Demographic characteristics of abusive (n = 80) and comparison (n = 86) mothers
African American 55 (69)61 (71)
Caucasian23 (29) 23 (27)
Other (Hispanic, biracial)2 (2) 2 (2)
Child sex 2.89 (1)
Male43 (54) 35 (41)
Female36 (46)50 (58)
I4 (5) 7 (9)
II 24 (32) 20 (28)
III 11 (15)17 (24)
IV14 (18) 12 (17)
V 23 (30)16 (22)
Marital status1.53 (3)
Single 31 (39) 34 (39)
Married/living together27 (34)34 (39)
Separated 10 (13)10 (12)
Divorced12 (15)8 (9)
Educational attainment7.08 (4)
Less than high school22 (28)12 (14)
High school graduate20 (25)20 (23)
Some college22 (28) 27 (31)
College graduate 14 (18)20 (23)
Graduate degree2 (3)7 (8)
Mean (SD)Mean (SD)
F value (df)
Parent age32.9 years (7.5)33.73 years (6.9) .48 (1,164)
Child age 7.4 years (1.7)7.0 years (1.3)3.29 (1,164)*
Parent IQ90.9 (14.3)94.7 (13.9)2.56 (1,145)
J Child Fam Stud
The Parenting Stress Index/Short Form (PSI/SF; Abidin 1995) is a 36-item self-
report measure of parenting stress. Parents use a 5-point scale to indicate the degree
to which they agree with each statement (e.g., ‘‘My child generally wakes up in a
bad mood’’; ‘‘I feel trapped in my responsibilities as a parent’’). This instrument
yields scores for several factors in addition to a Total Stress score, which was
utilized for purposes of the current research. PSI/SF scores are stable over time, are
internally consistent, and are significantly related to generalized distress (Reitman
et al. 2002) and observed parent behavior (Haskett et al. 2006). Internal consistency
of the Total scale for the larger study on which this sample was drawn was .83.
Observed Child Behavior
Each mother–child dyad participated in a 30-min interaction session videotaped for
later coding. During the first 10-min segment, dyads played together in a room with
a standard set of age-appropriate toys (i.e., building materials, markers and paper for
drawing, magnetic number and letter set). In the second 10-min segment mothers
were told to ask their child to clean up the toys, draw a picture of a person, and then
sit quietly while the mother read a magazine. In the final 10-min segment the mother
was instructed to help the child quickly complete two puzzles but she was told
not to touch the puzzle pieces. Only data from the first two segments were used in
the current study because data were missing from the third segment for several
participants due to time constraints in the data collection session. Child behavior
was coded from videotapes using a 7-point scale for Positive Mood, Negative
Mood, Persistence, and Compliance (adapted from Cox 1997). Positive Mood was
indicated by signs of contentment such as smiles, laughter, conversation with the
parent, and animated play. Negative Mood was indicated by signs of discontent
such as irritability, crying, and turning away from the parent. Persistence referred to
on-task behavior; children who scored high were consistently focused on the play
materials and/or their parent. Compliance referred to the degree to which the child
followed parental requests or commands.
Scoring of the scales involved ratings from 1 to 7 (with 7 indicating that the
category was highly indicative of the observed child behavior) on each of the four
child behaviors for each 10-min segment. Coders, naı ¨ve to the purpose of the study
and abuse status of mothers, were trained to 80% reliability using practice tapes of
parent–child interactions. Training required approximately 20 h. For more infor-
mation on training coders, see Paley et al. (2001). Reliability across coders was
examined for 34% of the interaction sessions (approximately equal numbers of
abusive and comparison mothers) based on independent coding, and kappa
coefficients ranged from .75 to .94 for exact agreement on codes. Bivariate
correlations of ratings across segments showed significant inter-segment consis-
tency in ratings, thus a mean score of 1–7 was generated by averaging ratings across
the two segments. Factor analysis supported further data reduction to a single scale
J Child Fam Stud
for child behavior. The scale does not include Positive Mood because scores for
that dimension of behavior did not load highly on the single factor (factor
loading = -.135). Negative Mood was reverse coded so that higher scores indicate
more positive child behavior. Internal consistency (Cronbach’s alpha) of the child
behavior scale was .80. There is abundant support for validity of this coding system.
Scores on the various scales used in this system have been shown to be related to
marital quality and children’s developmental status (see Paley et al. 2001), and have
been used to derive clinically-relevant subgroups of abusive parents (Haskett et al.
Maternal Intolerance of Child Conduct Problems
The Eyberg Child Behavior Inventory (Eyberg and Pincus 1999) was used to assess
parents’ intolerance for child behavior problems. The ECBI is a parent rating scale
comprised of 36 common child behavior problems. Parents first estimate (using a 7-
point scale) the frequency with which their child exhibits each of the behaviors
(Intensity Scale) and then they indicate whether (yes/no) each of the behaviors are
problematic for the parent (Problem Scale). The Problem scale was employed in the
current study; higher scores indicate greater intolerance (i.e., more child behaviors
are considered problematic). Raw scores of 15 on that scale are considered to be in
the clinical range. The ECBI is a widely used measure of behavior problems, is
sensitive to effects of parent training interventions, and has strong psychometric
properties. The authors report that the 12-week and 10-month test–retest stabilities
of Problem scores are high (.85 and .75, respectively) and the correlation between
mothers’ and fathers’ Problem scores is moderate (.61). ECBI scores are correlated
with observational measures of child affect and behavior during mother–child
interactions and with parent-reported temperament (see Eyberg and Pincus 1999).
For the larger data set on which this sample was drawn, internal consistency (Kuder-
Richardson-20) of the Problems scale was .93.
The Symptom Checklist 90-Revised (SCL-90-R; Derogatis 1983) is a self-report
inventory designed to assess patterns of psychological symptoms experienced by
adults. It consists of 90 symptoms that are rated on a 5-point scale to indicate
how much each symptom has bothered parents in the last 7 days. The Global
Severity Index (GSI), utilized in this study, combines the number of symptoms
reported and intensity of distress to yield a single indicator of current emotional
health. Raw scores were converted to T-scores using adult non-patient norms by
gender. Psychometric properties of the SCL-90-R have been examined exten-
sively (see Derogatis and Lazarus 1994). Specifically, convergent-discriminant
validity studies indicate SCL-90-R scales have moderate to high correlations
with corresponding MMPI scales. The SCL-90-R discriminates diagnostic
categories of psychopathology and is sensitive to changes in distress following
J Child Fam Stud
To aid in interpretation of findings, descriptive data and intercorrelations for all
study variables are presented in Tables 2 and 3, respectively. As shown in Table 2,
there were significant group differences in mothers’ parenting stress and psycho-
logical functioning and in children’s behavior. Differences in mothers’ level of
intolerance approached significance, with abusive mothers’ intolerance being
higher. Tests of primary hypotheses were based on regression analyses (Table 4).
In the first step, the three predictors and abuse status were entered. Together, the set
of predictors explained 40.3% of the variance in parenting stress, F(4, 113) = 19.1,
p\.01). The effect of psychological distress was significant, but maternal
intolerance and child behavior were not significant predictors of parenting stress
in the regression model. In the second step, interaction terms for each of the
predictors with abuse status were added to the model. The addition of the interaction
terms revealed an interesting pattern. Together, the interaction terms explained an
additional 1.84% of the variance in parental stress (DR2= .0184, p\.05). There
was not a significant interaction between psychological functioning and abuse
status or between child behavior and abuse status. However, there was a significant
interaction of maternal intolerance and abuse status in prediction of parenting stress.
To explore the nature of this interaction, a simple slopes analysis (Preacher et al.
Table 2 Scores on all measures for abusive and comparison mothers and full sample
(N = 166)
(n = 80)
(n = 86)
Child positive behaviora
6.15 (.72) 5.97 (0.76)6.31 (0.64)8.83.003
10.44 (8.59)11.75 (9.0)8.89 (7.7)3.46.06
55.97 (10.6)58.74 (10.7)53.36 (9.9) 11.13.001
82.99 (22.5)89.51 (23.0)77.23 (20.6) 12.83.0001
Note: t Values are for group differences between abusive and nonabusive mothers. N’s vary from 124 to
166 due to missing data on some measures
aRaw mean scores.bT-scores.cPercentile scores
Table 3 Intercorrelations among all variables
Note: Correlations above the axis are for abusive mothers. Those below the axis are for comparison
mothers. Ns vary from 53 to 86 due to missing data on some measures
* p\.05, ** p\.01, *** p\.001
J Child Fam Stud
2006) was conducted; this analysis produced a significance test of the regression
weight for each group independently. As seen in Fig. 1, the relation between
maternal intolerance and parenting stress was not significant for nonabused mothers.
However, for abused mothers, the relation between intolerance and parenting stress
was significant; greater intolerance for child misbehavior was associated with higher
levels of parenting stress for those mothers.
To address the question of the relative association between parenting stress and
child behavior versus maternal intolerance, analyses were conducted to compare the
correlation between parenting stress and child behavior (r = -.23) and the
correlation between parenting stress and intolerance (r = .55). The Fisher Z
transformation (c.f. Cohen et al. 2003) was used to convert the correlations into z
scores, allowing for the calculation of a 95% confidence interval for the difference
Table 4 Summary of hierarchical regression analysis for variables and interactions predicting parental
stress (N = 113)
Psychological distress 220.127.116.11**
Child positive behavior-3.102.37-0.09
Abuse status-3.21 3.46-0.07
Maternal intolerance0.652.32 0.03
Psychological distress1.18 0.210.53**
Child positive behavior-4.573.02-0.15
Child positive behavior 9 abuse status4.52 4.82 0.64
Psy. distress 9 abuse status-0.070.35-0.08
Parental intolerance 9 abuse status8.80 4.070.58*
Note: R2= .403 for Step 1; DR2= 1.84 for Step 2
* Significant at the .05 level; ** significant at the .01 level
Fig. 1 Interaction of maternal intolerance and abuse status in prediction of parenting stress. Note: Values
for low and high parenting stress are ±1 SD from the sample mean. Y-axis values are raw scores
J Child Fam Stud
between the two correlations. The resulting interval (.70, .93) indicated that the two
correlations were indeed significantly different from one another.
This investigation added to existing knowledge of the determinants of parenting
stress and explored the contributions of each predictor to parenting stress as
moderated by abuse status. The model, which included measures of children’s
behavior, mothers’ psychological distress, and maternal intolerance for child
conduct problems, accounted for a substantial portion of the variance in parenting
stress. Consistent with our expectations and with prior research, the determinants
included in our model provided a very good explanation of sources of stress in the
parenting role. By far, however, general psychological functioning was the single
best predictor of parenting stress. In correlational analyses, maternal intolerance for
child misbehavior was significantly related to parenting stress; however, when
entered in the regression model with general psychological distress, this predictor
did not make a contribution to the explanation of variance in parenting stress levels.
Nor did observed child behavior significantly predict parenting stress when included
in the model with psychological distress.
A probe of the significant interaction of maternal intolerance for child conduct
problems and abuse status showed that intolerance predicted parenting stress only
for the abusive mothers. Thus, abusive mothers were not only more intolerant of
child misbehavior (see Table 2) but their experience of stress in the parenting role
was affected by that intolerance. A secondary research question was whether
maternal intolerance or child behavior would be more closely associated with
parenting stress for abusive mothers. As expected, abusive mothers’ parenting stress
level was more closely related to their intolerance for child conduct problems than
to their children’s behavior during parent–child interactions. These results are
consistent with Mash and Johnston’s (1990) belief that parenting stress of abusive
parents resulted from their perceptions of the difficulty of their children’s behavior
rather than the children’s actual behavioral characteristics. Our findings are also
consistent with those of Bigras et al. (1996) and Creasey and Jarvis (1994)
indicating that parents’ cognitions related to child behavior might be more critical
determinants of stress in the parenting role than actual child adjustment. This
conclusion must be interpreted in light of the fact that our measure of child behavior
was based on a relatively brief 20-min interaction, which might have attenuated the
relation between child behavior and parenting stress. Furthermore, shared method
variance associated with our measures of stress and intolerance (i.e., both constructs
were measured by parent report) likely maximized the relation between those
Some investigators have found that physically abusive mothers’ reports of their
children’s behavior are more negative than those of nonabusive parents even when
observers perceive no differences in behavior of abused and nonabused children.
Such findings have been interpreted as a negative bias among abusive parents
(e.g., Mash and Johnston 1990; Reid et al. 1987). This pattern of exaggeration does
J Child Fam Stud
not appear to characterize our particular sample of abusive mothers. Abused
children in this sample did in fact express a more negative mood and were less
compliant and persistent during play with their mothers. Although we did not
directly assess the accuracy of mothers’ perceptions of their children’s adjustment,
our findings challenge the notion that abusive mothers’ reports of their children’s
problems are based solely on a negative perceptual bias.
In conclusion, mothers’ level of stress in the parenting role was largely accounted
for by their general psychological functioning. This was true for both abusive and
nonabusive mothers. Thus, approaches to decrease mothers’ experiences of pressure
and strain as they raise young children could be similar for all mothers, regardless of
abuse status. Specifically, efforts to improve general psychological well-being of
mothers might be useful in protecting against parenting stress, as suggested by
Webster-Stratton (1990). However, findings also suggest that it might be important
to provide an additional intervention component for abusive parents to focus on the
impact of children’s misbehavior on parents’ stress levels. Such efforts could
include strategies to modify mothers’ interpretations and perceptions of their
children’s misbehavior. Findings support intervention practices for physically
abusive parents that include identifying and directly challenging cognitive
distortions and negative scripts related to children’s problematic behaviors (Kolko
and Swenson 2002).
A strength of this study is the use of observational data to measure children’s
adjustment during interactions with their mothers; however, it is possible that
observations in the home environment might have produced variables more strongly
related parenting stress. Further, psychological functioning as well as parenting
stress were measured via mothers’ self-report, which likely increased shared method
variance. Furthermore, the two measures of distress likely tapped into similar
constructs due to some degree of overlap in item content. Interview measures of
psychological functioning could be used in future investigations to possibly reduce
the threat of shared method variance.
Although information related to severity of abuse was not available for this
sample, children resided with their parents at the time of data collection so it is
possible that this sample experienced mild to moderate abusive parenting rather than
severe abuse. It is also possible that some of the families received services between
the time of the abuse report and participation in this study. Thus, replication will be
important to establish whether these findings generalize to the full range of abusive
parenting. Unfortunately we lacked a sufficient sample of fathers to examine our
models separately by parent gender. That is regrettable because many questions
related to gender differences in determinants of parenting stress remain unanswered
(Creasey and Reese 1996; Deater-Deckard 1998). Such research might be
informative in further refining models of parenting stress and modifying interven-
tion approaches so they are sensitive to the needs of mothers and fathers.
Given the wide range of sources of parenting stress and potential factors that
protect against high levels of stress, our model was somewhat limited in scope.
The model did account for a large proportion of the variance in parenting stress,
however, there was unexplained variance remaining that could be captured by
expanded models in future research. For example, an expanded model could include
J Child Fam Stud
parenting style as a predictor of children’s behavior, which in turn predicts
parenting stress. In addition, the presence or absence of social support might be a
critical factor to consider in understanding parenting stress (Budd et al. 2006; Bonds
et al. 2002).
It is impossible to determine, on the basis of this correlational research, the
direction of causality among the variables in our model. Level of parenting stress
might indirectly impact child behavior, for example, through the negative effects of
stress on parenting practices (Deater-Deckard and Scarr 1996).Longitudinal studies
will be needed to identify the processes involved in linkages among and between
risk and protective factors for parenting stress. Finally, the focus of this
investigation was determinants of parenting stress. Future studies should be
designed to test whether outcomes of parenting stress differ for abusive and
nonabusive parents. Even though correlates appear to be similar for these two
groups, parenting stress might be associated with different parenting behaviors and/
or child outcomes for abusive and nonabusive parents.
awarded to Haskett. The authors express gratitude to the families who participated in the Parents and
Children Together Project.
This research was supported by a grant from the National Institute of Mental Health
Abidin, R. R. (1992). The determinants of parenting behavior. Journal of Clinical Child Psychology, 21,
Abidin, R. R. (1995). Parenting stress index professional manual (3rd ed.). Odessa, FL: Psychological
Assessment Resources, Inc.
Azar, S. T. (1986). A framework for understanding child maltreatment: An integration of cognitive
behavior and developmental perspectives. Canadian Journal of Behavioural Science, 18, 340–355.
Bigras, M., LaFreniere, P. J., & Dumas, J. E. (1996). Discriminant validity of the parent and child scales
of the Parenting Stress Index. Early Education and Development, 7, 157–178.
Bolger, K., & Patterson, C. (2001). Pathways from child maltreatment to internalizing problems:
Perceptions of control as mediators and moderators. Development and Psychopathology, 13, 913–
Bonds, D. D., Gondoli, D. M., Sturge-Apple, M. L., & Salem, L. N. (2002). Parenting stress as a mediator
of the relation between parenting support and optimal parenting. Parenting: Science and Practice, 2,
Bradley, E. J., & Peters, R. D. (1991). Physically abusive and nonabusive mothers’ perceptions of
parenting and child behavior. American Journal of Orthopsychiatry, 61, 455–460.
Budd, K. S., Holdsworth, M. J. A., & HoganBruen, K. D. (2006). Antecedents and concomitants of
parenting stress in adolescent mothers in foster care. Child Abuse & Neglect, 30, 557–574.
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation analysis
for the behavior sciences (3rd ed.). Mahway, NJ: LEA Publishers.
Cox, M. (1997). Qualitative ratings of parent/child interaction. Unpublished manual, University of North
Carolina at Chapel Hill.
Creasey, G., & Jarvis, P. A. (1994). Relationships between parenting stress and developmental
functioning among 2-year olds. Infant Behavior and Development, 17, 423–429.
Creasey, G., & Reese, M. (1996). Mothers’ and fathers’ perceptions of parenting hassles: Associations
with psychological symptoms, nonparenting hassles, and child behavior problems. Journal of
Applied Developmental Psychology, 17, 393–406.
Deater-Deckard, K. (1998). Parenting stress and child adjustment: Some old hypotheses and new
questions. Clinical Psychological Science and Practice, 5, 314–332.
Deater-Deckard, K. (2004). Parenting stress. New Haven, CT: Yale University Press.
J Child Fam Stud
Deater-Deckard, K., & Scarr, S. (1996). Parenting stress among dual-earner mothers and fathers: Are
there gender differences? Journal of Family Psychology, 10, 45–59.
Derogatis, L. R. (1983). SCL-90-Revised: Administration, scoring, and procedures manual - II. Towson,
MD: Clinical Psychometric Research.
Derogatis, L. R., & Lazarus, L. (1994). SCL-90-R, Brief Symptom Inventory, and matching clinical
rating scales. In M. E. Maruish (Ed.), The use of psychological testing for treatment planning and
outcome assessment (pp. 217–248). Hillsdale, NJ: LEA, Inc.
Ethier, L. S., Lacharite, C., & Couture, G. (1995). Childhood adversity, parental stress, and depression of
negligent mothers. Child Abuse & Neglect, 19, 619–632.
Eyberg, S. M., & Pincus, D. (1999). Eyberg child behavior inventory professional manual. Psychological
Assessment Resources, Inc.: Odessa, FL.
Fischer, M. (1990). Parenting stress and the child with attention deficit hyperactivity disorder. Journal of
Clinical Child Psychology, 19, 337–346.
Gotlib, I. H., & Hanunen, C. L. (1992). Psychological aspects of depression: Toward a cognitive
interpersonal integration. Toronto, Canada: Wiley Series in Clinical Psychology.
Haskett, M. E., Ahern, L. S., Sabourin Ward, C., & Allaire, J. C. (2006). Factor structure and validity of
the Parenting Stress Index/Short Form. Journal of Clinical Child and Adolescent Psychology, 35,
Haskett, M. E., Smith Scott, S., & Sabourin Ward, C. (2004). Subgroups of physically abusive parents
based on cluster analysis of parenting behavior and affect. American Journal of Orthopsychiatry, 74,
Holden, E. W., & Banez, G. A. (1996). Child abuse potential and parenting stress within maltreating
families. Journal of Family Violence, 11, 1–12.
Kolko, D. J., & Swenson, C. C. (2002). Assessing and treating physically abused children and their
families: A cognitive-behavioral approach. Thousand Oaks: Sage Publications.
Mash, E., & Johnston, C. (1990). Determinants of parenting stress: Illustrations from families of
hyperactive children and families or physically abused children. Journal of Clinical Child
Psychology, 19, 313–328.
Mash, E., Johnston, C., & Kovitz, K. (1983). A comparison of the mother–child interactions of physically
abused and non-abused children during play and task situations. Journal of Clinical Child
Psychology, 12, 337–346.
Milner, J. S. (2003). Social information processing in high-risk and physically abusive parents. Child
Abuse & Neglect, 27, 7–20.
Paley, B., Cox, M. J., & Kanoy, K. W. (2001). The young family interaction coding system. In P. K.
Kerig & K. M. Lindahl (Eds.), Family observational coding systems: Resources for systematic
research (pp. 273–288). Mahwah, NJ: Lawrence Erlbaum Associates.
Pinderhughes, E. S., Dodge, K. A., Bates, J. E., Pettit, G. S., & Zelli, A. (2000). Discipline responses:
Influences of parents’ socioeconomic status, ethnicity, beliefs about parenting, and cognitive-
emotional processes. Journal of Family Psychology, 14, 380–400.
Preacher, K. J., Curran, P. J., & Bauer, D. J. (2006). Computational tools for probing interaction effects in
multiple linear regression, multilevel modeling, and latent curve analysis. Journal of Educational
and Behavioral Statistics, 31, 437–448.
Quittner, A., Gluekauf, R., & Jackson, D. (1990). Chronic parenting stress: Moderating versus mediating
effects of social support. Journal of Personality and Social Psychology, 59, 1266–1278.
Reid, J., Kavanagh, K., & Baldwin, D. (1987). Abusive parents’ perceptions of child problem behaviors:
An example of parental bias. Journal of Abnormal Child Psychology, 15, 457–466.
Reitman, D., Currier, R. O., & Stickle, T. R. (2002). A critical evaluation of the Parenting Stress Index-
Short Form (PSI-SF) in a Head Start population. Journal of Clinical Child and Adolescent
Psychology, 31, 384–392.
Rodriguez, C., & Green, A. (1997). Parenting stress and anger expression as predictors of child abuse
potential. Child Abuse & Neglect, 21, 367–377.
Rogers, A. Y. (1993). The assessment of variables related to the parenting behaviors of mothers with
young children. Child and Youth Services Review, 15, 385–402.
Ross, C. N., Blanc, H. M., McNeil, C. B., Eyberg, S. M., & Hembree-Kigin, T. L. (1998). Parenting stress
in mothers of young children with oppositional defiant disorder and other severe behavior problems.
Child Study Journal, 28, 93–110.
Schellenbach, C., Monroe, L., & Merluzzi, T. (1991). The impact of stress on cognitive components of
child abuse potential. Journal of Family Violence, 6, 61–80.
J Child Fam Stud
Straus, M. A. (1990). Manual for the Conflict Tactics Scale. Durham, NH: Family Research Laboratory, Download full-text
University of New Hampshire.
Straus, M. A., & Hamby, S. L. (1997). Measuring physical and psychological maltreatment of children
with the Conflict Tactics Scales. In G. K. Kantor & J. L. Jasinski (Eds.), Out of darkness:
Contemporary perspectives on family violence (pp. 119–135). Thousand Oaks, CA: Sage
Webster-Stratton, C. (1990). Stress: A potential disruptor of parent perceptions and family interactions.
Journal of Clinical Child Psychology, 19, 302–312.
Williford, A. P., Calkins, S. D., & Keane, S. P. (2007). Predicting change in parenting stress across
childhood: Child and maternal factors. Journal of Abnormal Child Psychology, 35, 251–263.
Wolfe, D. A. (1999). Child abuse: Implications for child development and psychopathology. Thousand
Oaks, CA: Sage Publications, Inc.
J Child Fam Stud