Emotional flooding and hostile discipline in the families of toddlers with disruptive behavior problems

Emotional Flooding and Hostile Discipline in the Families of ToddlersWith Disruptive Behavior Problems

Melanie Mence, David J. Hawes,and Lucinda Wedgwood

University of Sydney

Susan Morgan, Bryanne Barnett, and Jane KohlhoffKaritane, Sydney South West Area Health Service, Villawood,

Australia

Caroline HuntUniversity of Sydney

This study examined the relationship between negative parenting practices and dysfunction in parents’cognitive processing of child affect cues in families of toddlers with disruptive behavior problems. Thisdysfunction comprised a bias toward the misclassification of child affect as anger (affect appraisal bias)and parents’ proneness to emotional flooding (Gottman, 1991, 1993). Participants were families oftoddlers (n � 82; 53% male; aged 18 – 48 months) referred to a tertiary-level health service for thetreatment of disruptive behavior problems. Affect appraisal bias was indexed in terms of the discrepancybetween rates of child anger coded from video recordings of parent–child interactions and rates of childanger estimated by parents immediately after these interactions. Parenting practices and emotionalflooding were assessed using the Parenting Scale and the Parental Flooding Scale. Both hostile andoverreactive discipline were positively associated with severity of disruptive behavior problems, howeveronly hostile discipline was associated with the biased appraisal of child affect and emotional flooding.Emotional flooding was found to be a unique predictor of hostile discipline, independent of covariatesincluding the severity of disruptive behavior problems. Variance in hostile discipline was furtherexplained by the interaction between emotional flooding and affect appraisal bias. Emotional floodingappears to be particularly proximal to hostile discipline in the families of toddlers with disruptivebehavior problems, consistent with evidence previously reported for nonclinical families.

Keywords: disruptive behavior problems, early childhood, parenting practices, emotion

It is now well recognized that the most chronic trajectories ofantisocial behavior are initiated in early childhood, and shapedby both biological and family-based mechanisms (Moffitt &Caspi, 2007; Siever, 2008). Causal models have long empha-sized the importance of negative parenting practices to risk forantisocial behavior, with extensive evidence associating earlyonset disruptive behavior problems (e.g., developmentally ex-cessive defiance and aggression) with high levels of harsh,overreactive, and inconsistent/lax discipline (see Hawes &Dadds, 2005). Much is known about the parent– child dynamicsthrough which such parenting practices maintain and amplifydisruptive behavior problems across the toddler and preschoolyears, with dominant conceptualizations emphasizing the pa-rental modeling of aggression, as well as escalating cycles ofcoercion based on escape-avoidance conditioning. In Patter-

son’s (1982) coercion model, these cycles are seen to functionas “reinforcement traps” that reward both parents’ and chil-dren’s use of aversive control tactics (e.g., whining, nagging,shouting, hitting), and extinguish positive family interactions.According to Patterson (2002), the coercive cycles that charac-terize the families of children with early onset antisocial be-havior are often established by the time such children are twoyears of age. It is also assumed that the upper amplitudes ofsuch cycles will be pushed higher and higher over time, withparental hostility becoming increasingly frequent and intense(Dishion & Patterson, 2006).

Individual differences in negative parenting have been associ-ated with a range of ecological and individual-level factors, in-cluding socioeconomic adversity, parental psychopathology, andpersonality variables (Dix, 1991; Vondra & Belsky, 1993). Impor-tantly, cognitive risk factors appear to be particularly proximal tonegative parenting behaviors, both in the families of children withdisruptive behavior problems as well as families at risk for childmaltreatment. Much research in this area has focused on parentalattributions concerning child behavior, with established findingsindicating that parents who engage in harsh and inconsistent dis-cipline often hold explanations for child difficulties that emphasizethe child’s hostile intent, negative characteristics of the parent, orhistorical experiences in the family (see review by Johnston &Ohan, 2005).

This article was published Online First January 6, 2014.Melanie Mence, David J. Hawes, and Lucinda Wedgwood, School of

Psychology, University of Sydney, Sydney, Australia; Susan Morgan,Bryanne Barnett, and Jane Kohlhoff, Karitane, Sydney South West AreaHealth Service, Villawood, Australia; Caroline Hunt, School of Psychol-ogy, University of Sydney.

Correspondence concerning this article should be addressed to David J.Hawes, School of Psychology, University of Sydney, Sydney Australia,NSW 2006. E-mail: [email protected]

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

Journal of Family Psychology © 2014 American Psychological Association2014, Vol. 28, No. 1, 12–21 0893-3200/14/$12.00 DOI: 10.1037/a0035352

12

In addition to parental attributional biases, there is growingevidence that parenting practices associated with disruptive behav-ior problems may be accounted for by social information process-ing biases concerning parents’ attention to (i.e., encoding) andinterpretation (i.e., appraisal) of child cues. Findings suggest thatthese processing biases may be of particular importance to explain-ing harsh parenting in the early childhood years (e.g., Lorber &O’Leary, 2005; Lorber, O’Leary, & Kendziora, 2003). One of themost intensive studies of these processes to date was conducted byLorber et al. (2003), using a video-mediated recall paradigm withmothers of toddlers (n � 40) aged 16–47 months. Mothers’negative appraisal bias was related to overreactive discipline,which is often defined in terms of excessive pickiness and esca-lating irritability (e.g., Rhoades & O’Leary, 2007). This effect wasindependent of levels of disruptive child behavior, and strongestamong parents who demonstrated a preferential negative encodingbias—that is, parents who preferentially attended to negative childbehavior over positive behavior. Likewise, Lorber and O’Leary(2005) found that such an appraisal bias accounted for the rela-tionship between parental physiological reactivity (i.e., angerarousal) and harsh discipline. Although often associated with theanger and irritability of overreactive discipline, harsh discipline istypically defined by a parents’ use of physical or verbal force (e.g.,Rhoades & O’Leary, 2007). These findings are consistent withmodels in the clinical and child maltreatment literature that assumethat parents who engage in harsh and overreactive discipline aremore likely to notice their children behaving badly than behavingwell, and are more likely than other parents to interpret neutral orpositive child behavior to be misbehavior (e.g., Milner, 1993;Wahler & Dumas, 1989; Patterson, 1982).

Although such research has focused primarily on biases as-sociated with the processing of disruptive child behavior cues,there is growing evidence to suggest that negative parentingpractices may implicate the biased processing of child affect. Ina study investigating the dynamic associations between moth-ers’ self-reported emotion and observed child cues, Lorber andSlep (2005) found that levels of harsh (as well as lax) parentingwere highest among mothers whose own emotions were highlydependent on their child’s negative affect. Conversely, no suchnegative parenting was exhibited by mothers whose emotionalreactions were highly dependent on their child’s rule violations.The authors concluded that parents’ processing of child affectmay have greater implications for their parenting practices thantheir processing of cues specific to oppositional/defiant behav-ior. Such evidence converges with findings from other lines ofresearch that have examined affect appraisal biases in parents.For example, in an observational study of family problem-solving discussions, parents of children with conduct disorderwere found to overreport child anger cues significantly morethan parents of nonproblem children (Sanders et al., 1992).Similar findings have been reported for parents of children withdisruptive behavior problems in experimental research involv-ing standardized emotional stimuli (Snarr et al., 2003). There isalso some evidence that biases in the recognition of child affectcues may characterize parents at risk for physical abuse, how-ever findings have been mixed (see Asla, de Paúl, & Pérez-Albéniz, 2011). Despite the apparent importance of this evi-dence to clinical models of disruptive behavior problems,associations between parenting practices and affect appraisal

biases have not previously been examined in the families ofchildren with clinically severe disruptive behavior problems.

Emotional Flooding and Negative Parenting

In clinical models of family dysfunction, risk mechanisms infamily members’ processing of each others’ affective and be-havioral cues have often been conceptualized in terms of ‘emo-tional flooding.’ Drawing on Ekman’s (1984) model of emo-tional conditioning, Gottman (1991) first applied the concept ofemotional flooding to family systems to explain relationshipdysfunction in distressed couples. Flooding, as operationalizedin the self-report measure used throughout Gottman’s programof research on relationship breakdown (see Gottman, 1993),occurs when an individual experiences another’s negative emo-tional cues as unexpected, unprovoked, intense, overwhelming,and disorganizing. According to Gottman (1993), emotionalflooding is central to relationship breakdown, and occurs whenone partner increasingly perceives the other’s negative affect asunpredictable and overwhelming. This is assumed to lead toemotional (escape) conditioning, involving a state of hypervigi-lance to the cues conditioned to the flooding. The floodedindividual then begins to distort ambiguous cues in the biaseddirection of seeing them as threatening or frustrating, which inturn makes the individual more prone to experience flooding. Adefining characteristic of emotional flooding is that it disruptshigher-order cognitive processes such as those involved inproblem-solving, thereby compromising the individual’s capac-ity to respond adaptively in conflict situations. Importantly, thisdefinition of emotional flooding emphasizes that the floodedpartner will do anything to terminate the partner interactionassociated with the sensation, typically engaging in overlearnedresponses that provide immediate escape (e.g., leaving theroom). For the most detailed process-level accounts of emo-tional flooding see Gottman (1991, 1993).

The concept of emotional flooding has subsequently been ap-plied to the coercive parent–child interactions in the families ofchildren with disruptive behavior problems. Snyder et al. (1994)proposed that Gottman’s conceptualization of flooding may rep-resent a mechanism through which child anger compromises par-ents’ capacity to effectively problem-solve discipline scenarios,resulting in a reliance on overlearned strategies. This is consistentwith research showing that maternal anger organizes and directsbehavior toward coercion and away from strategies that soothe andreduce child arousal (Dix, 1991). Just as emotional flooding isassociated with attempts to terminate distressing partner interac-tions in discordant couples, it is assumed to be associated withattempts by flooded parents to terminate the aversive parent–childinteractions. However, whereas escape from a partners’ distressingemotional cues often involves literally escaping from the scene ofthe interaction, this strategy is less likely to be available to floodedparents of young children. Rather, aversive parent–child interac-tions often escalate until terminated by a sharp escalation in parentcoercion in the form of an explosive verbal/physical reaction suchas a scream or hit (Patterson, 1982). It has been proposed thatrepeated episodes of such escalating mutual coercion and angermay interfere with the construction of an interaction pattern con-ducive to the child’s acquisition of self-regulation (Snyder et al.,1994).

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

13EMOTIONAL FLOODING AND HOSTILE DISCIPLINE

The association between emotional flooding and parentingpractices has been explicitly tested in only two studies to date.Slep and O’Leary (2007) first examined the relationship be-tween emotional flooding and parental aggression in a commu-nity sample of families of children aged 3–7 years, using aself-report measure of the construct that indexes the extent towhich children’s negative affect is experienced by the parent tobe overwhelming and unpredictable, and interferes with theparent’s capacity to think clearly (Slep & Heyman, 1998). Datafrom this cross-sectional study showed not only that emotionalflooding predicted high levels of overreactive discipline andparent-to-child aggression, but that this association was inde-pendent of effects for conceptually related factors such ashostile attributions for child behavior, parental depression, andparental anger levels (Slep & O’Leary, 2007). In a subsequentstudy involving the same data set, Lorber et al. (2011) showedthat the relationship between flooding and negative parentingdoes not vary across race. Slep and O’Leary (2007) concludedthat ongoing research is needed to understand the unique rolethat emotional flooding appears to play in contributing to neg-ative parenting practices, given its potential as a target inparenting interventions for disruptive behavior problems.

Study Aims and Hypotheses

The aim of this study was to examine the relationship be-tween negative parenting practices and dysfunction in parents’processing of child affect in families of toddlers referred for thetreatment of disruptive behavior problems. Two forms of affectprocessing dysfunction were investigated. The first was a biastoward the misclassification of child affect as anger— herebyreferred to as affect appraisal bias. This was operationalized interms of the discrepancy between rates of child anger codedfrom video recordings of parent– child interactions and rates ofchild anger estimated by parents immediately after these inter-actions. Consistent with previous research into emotion-relateddynamics of parenting in the families of children with disrup-tive behavior problems, we chose to focus specifically on anger,because it is the emotion that most commonly characterizesdisruptive behavior in the toddler years (Cole, Teti, & Zahn-Waxler, 2003). Furthermore, there is research to suggest that itis the emotion that is most often associated with appraisalbiases in parents of children with conduct disorder (Sanders etal., 1992). The second form of affect processing dysfunctionexamined was emotional flooding, as operationalized using theself-report measure of the construct that has been used inresearch with families of typically developing children (Slep &O’Leary, 2007; Lorber et al., 2011). We are, however, aware ofno previous studies in which emotional flooding has beenexamined in the families of children with clinically significantbehavior problems.

Based on Gottman’s (1991, 1993) conceptualization of emo-tional flooding, it was predicted first that parents who wereprone to emotional flooding would exhibit a stronger affectappraisal bias during interactions with their children. Second, itwas predicted that both affect appraisal bias and emotionalflooding would be positively associated with negative parentingpractices, independent of the severity of disruptive behaviorproblems in the family. Third, theoretical accounts of emotional

flooding and family dysfunction emphasize an interplay be-tween emotional arousal and cognitive mechanisms. These con-ceptualizations place a primary emphasis on the physiologicalarousal and associated disruption to higher order cognitiveprocesses that characterize emotional flooding. Additionally, itis thought that risk for coercive family dynamics is greatestwhen individuals who are prone to emotional flooding developa biased interpretation of the affective cues of family members(Gottman, 1991, 1993). As such, the combination of emotionalflooding with a biased interpretation of affect is thought toconfer a stronger risk for negative parenting practices thaneither of these components in isolation. Based on this theory,we tested the hypothesis that negative parenting practices wouldbe predicted by an interaction between emotional flooding anda biased negative appraisal of child affect.

Method

Participants

Participants were families of toddlers (n � 82; 53.7% male;aged 18 – 48 months) referred to a tertiary-level community-basedearly childhood health service in Sydney, Australia, that special-izes in the treatment of disruptive behavior disorders in the toddleryears. This age range is comparable with previous studies ofcognitive mechanisms associated with negative parenting practices(e.g., Smith & O’Leary, 1995), with developmental evidence in-dicating that negative child affect in the context of disciplineoccurs at a particularly high rate in this age group (Roberts &Strayer, 1987). Children with developmental delays were excludedfrom the sample, as were parents with current substance abuseproblems or symptoms of psychosis. Parents were required to befluent in English.

Approximately 50% of the families who met the inclusioncriteria provided consent to participate, with the most commonreason for declining being the wish not to be videotaped. Themajority of children in the sample (75.6%) were the first born intheir family, and most (67.1%) had one sibling. The average agesof mothers and fathers were 32 years (SD � 4.7, range � 19–44years) and 34 years (SD � 5.3, range � 19–54 years), respec-tively. Most parents were married or in a defacto relationship(85.4%), whereas 14.6% considered themselves to be single par-ents. In the majority of cases (97.6%) the mother was identified asthe primary caregiver, and was the parent who completed the studymeasures. All children were born in Australia, whereas 28% ofmothers and 39% of fathers were born elsewhere, most oftenMiddle Eastern, European, and South East Asian countries. Themajority of mothers (65.6%) and fathers (93.9%) were employedeither full/part-time. Occupational status was coded using theAustralian Standard Classification of Occupations (Australian Bu-reau of Statistics, 1997), with the distribution of parent occupationin the sample being largely representative of that for the populationof Sydney (Australian Bureau of Statistics, 2007).

Procedure

Ethics approval for this study was obtained from the SydneySouth West Area Health Service Human Research Ethics Com-mittee – Western Zone. Families meeting the inclusion criteria

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

14 MENCE ET AL.

were identified at the point of telephone intake for new referrals tothe service. Parents were provided with information about theresearch in the initial clinical interview. Participating parents pro-vided written consent after this interview as part of a more detaileddiscussion of the project with a researcher who was not involvedin the clinical care of the family, so as to avoid any inadvertent orperceived coercion. All self-report and observational measureswere completed in the fortnight that followed as part of thefamily’s pretreatment assessment.

Measures

Affect appraisal bias. The measurement of parents’ biasedappraisal of child affect involved a number of observational andself-report procedures. First, parent–child interactions were vid-eotaped during a structured observation involving three compo-nents, each of which lasted approximately five minutes. Thesewere (1) child-led play, in which the parent was instructed to allowthe child to lead the play as the parent follows; (2) parent-led play,in which the parent chose an activity and took charge of the play;and (3) clean-up, in which the parent directed the child to clean upthe play things without assistance. This three-part structure wasdesigned to vary the degree of control required by the parent, andin turn elicit the child’s typical patterns of compliance versusoppositional and disruptive behavior across various contexts. It hasbeen used extensively in observational research into oppositionaland aggressive behavior in young children, and is regarded as aneffective observational paradigm for children up to the age of eightor so years (Hawes & Dadds, in press).

Second, immediately after the period of observation, parentscompleted a rating scale that assessed their subjective appraisals oftheir child’s affect during the preceding play interaction. Thisglobal rating required parents to estimate the proportion of theobservation period during which their child was feeling angry.Parents responded to the question “For how much time duringtoday’s session do you feel that your child was feeling angry/frustrated?”, on an 11-point scale that ranged from 0 (none of thetime) to 10 (100% of the time).

Third, independent observers coded rates of child anger fromvideo recordings of the parent–child observations using the Spe-cific Affect (SPAFF) coding system (Coan & Gottman, 2007). TheSPAFF is a well-validated system for coding displays of discreteemotions based on a gestalt of facial, vocal, and physical featuresof behavior, and has been used to code affect in previous obser-vational studies of child anger and parenting practices (e.g., Snyderet al., 2010). The current study used the anger code from thissystem, which was recorded using a partial interval time-samplingmethod in which the presence/absence of any displays of angerwere coded for each 15-s interval of the observation period. Ratesof anger were then computed for each case by dividing the numberof anger intervals by the overall number of intervals for which rawobservational data were collected. Training involved coding tohigh criterion agreement, and weekly recalibration and codingtraining sessions were used to minimize drift. Approximately 25%of all observations were used to assess coder reliability, and coderswere unaware of which observation sessions were to be used forreliability assessment. An intraclass correlation of .95 indicated ahigh level of reliability between coders. This microsocial codingwas used to collect the most reliable observational data on child

affect, and was therefore more intensive than the global ratings thatwere feasible for parents to complete in the setting in which theresearch was conducted. This was considered appropriate, giventhat we were interested not in comparing ratings of parents versusindependent observers but rather in individual differences in par-ents’ deviations from independent observers.

An index of affect appraisal bias was then calculated by com-puting the difference between parents’ ratings of their child’saffect and independent observers’ ratings of the child’s affect.Given that a different metric was used for parents’ ratings of childaffect (0–10) and independent observers’ coding of child affect(0–100), these were standardized before computing the differencescore for each case. Rates of child anger coded from video record-ings of the play task by independent observers were then sub-tracted from the rates of child anger estimated by parents imme-diately after the play task. As such, a difference score with a higherpositive value indicated a greater bias toward misclassifying childcues as expressions of anger. Similar approaches to measuringnegative bias in parents’ appraisals of child affect and behaviorhave proven to be informative in clinical research (see Hawes,Dadds & Pasalich, in press). An example closely related to thecurrent study is that of Sanders et al. (1992), in which bias wasindexed in terms of the discrepancy between observations of childaffect coded from video recordings of parent–child problem-solving discussions, and mothers’ reports of their children’s affectduring these problem-solving discussions, which mothers providedimmediately after the discussions.

Parenting practices. The Parenting Scale (PS; Arnold et al.,1993) is 30-item self-report measure of dysfunctional disciplinepractices in parents of young children over the preceding twomonths. Each item pairs a dysfunctional discipline practice with itsmore effective counterpart to form the anchors for a 7-point scale(e.g., When my child misbehaves: I raise my voice or yell/I speakto my child calmly). The validity and reliability of the PS hasreceived much empirical support in both community and clinicalsamples (Duppong-Hurley et al., 2013). Measurement research hasindicated support for various approaches to the scoring of themeasure, with the current study using the three-factor model re-ported by Rhoades and O’Leary (2007), based on confirmatoryfactor analysis of the solutions reported in five prior studies of thePS in families with young children. This method for scoring themeasure produces subscales for discipline that is lax (e.g., Whenmy child does something I don’t like, I often let it go), overreactive(When I’m upset or under stress, I am picky and on my child’sback), and hostile (e.g., When my child misbehaves, I spank, slap,grab, or hit my child, When my child misbehaves, I almost alwaysuse bad language). These subscales showed adequate internalconsistency in the current sample (�s � .63 – .68).

Emotional flooding. The Parental Flooding Scale (PFS; Slep& Heyman, 1998) was designed to index Gottman’s conceptual-ization of emotional flooding, as applied to parents’ responses tochild affect. As such, items assess the degree to which parents feelthat their child’s negative affect is unpredictable and overwhelm-ing, and is associated with disruptions to higher order cognitiveprocessing (e.g., My child tends to explode without any warningsigns, The intensity of my child’s distress catches me off guard, Iget all jumbled when my child is upset, I can’t think straight whenmy child is upset with me). Items are rated on a scale ranging from1 (almost always) to 5 (never). The measure was originally de-

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

15EMOTIONAL FLOODING AND HOSTILE DISCIPLINE

signed for use in the study of predictors of parental aggressionreported by Slep and O’Leary (2007). Psychometric evaluation ofthe measure conducted in the course of that study supported ascoring method in which nine of the original 15 items are retained,producing a scale with very good internal consistency (� � .93;Slep & O’Leary, 2007). This same 9-item scale was adopted foruse in the current study, in which evidence of very good internalconsistency was also found (� � .86).

Child disruptive behavior problems. The Eyberg Child Be-havior Inventory (ECBI; Eyberg & Pincus, 1999) is a 36-itemparent-report measure of the occurrence of disruptive behaviorproblems in children. The ECBI comprises an Intensity Scale thatmeasures the frequency of problem behaviors on a 7-point scale(ranging from never to always) and a Problem Scale, which asksparents to report whether the behavior is perceived to be a problem(yes or no). Extensive support for the reliability and validity ofthese two scales has been reported previously (e.g., Boggs et al.,1990; Eyberg & Pincus, 1999; Rich & Eyberg, 2001). Excellentinternal consistency was seen for both the Intensity (� � .94) andProblem (� � .88) scales in the current sample. The mean scorefor the sample on the ECBI Intensity scale (M � 128.84; SD �36.99) exceeded the score of 127 commonly used as a clinicalcutpoint for the measure (Burns & Patterson, 1990). Although themajority of participants (54%) exceeded this clinically severethreshold, a large proportion of the sample (46%) were referred fordifficulties managing somewhat less severe problems associatedwith disruptive behavior.

Results

Descriptive statistics and bivariate correlations among the keystudy variables are reported in Table 1. Observational variablesderived from the parent–child interaction task were log trans-formed because of skewness before analysis. Parents’ subjectiveratings of child anger cues immediately after play interactionscorrelated moderately with rates of observed child anger codedfrom video recordings of these interactions by independent observ-ers (r � .47, p � .00). However, mean rates of child angerestimated by parents (M � 14.25, SD � 17.41) were significantlyhigher than those coded by independent observers (M � 3.00,SD � 5.22; t(79) � �6.53, p � .001), with those rates estimated

by parents exceeding those coded by observers in 88.75% of cases.Rates of child anger coded by independent observers ranged from0–28%, whereas rates estimated by parents ranged from 0% to90%.

The index of affect appraisal bias was significantly associatedwith emotional flooding (r � .23, p � .05), and hostile discipline(r � .24, p � .05). This index was also significantly associatedwith severity of disruptive behavior problems as indexed by ECBIProblem (r � .25, p � .05) and Intensity (r � .23, p � .05) scores.Among the negative parenting practices measured, hostile disci-pline was associated with parental flooding (r � .39, p � .01) andseverity of disruptive behavior problems (ECBI Intensity: r � .24,p � .05). Overreactive discipline was positively associated withseverity of disruptive behavior problems (ECBI Problem: r � .29,p � .05; Intensity .30, p � .01), whereas lax discipline was notassociated with any variables aside from other negative parentingpractices. Child age was associated with conduct problem severity(ECBI Problem: r � .46, p � .05; Intensity r � .42, p � .01).Neither mother nor father age were associated with any studyvariables other than each other (r � .64, p � .001). Single-parentstatus was significantly associated with higher levels of hostilediscipline t(77) � �2.08, p � .05) and emotional floodingt(77) � �2.78, p � .01). Socioeconomic status was significantlyassociated with affect appraisal bias (r � �.25, p � .05), emo-tional flooding (r � �.29, p � .05), and ECBI Problem scores(r � �.29, p � .05). Socioeconomic status did not significantlycorrelate with any other variables, and no sex differences wereindicated for any of the study variables (p � .19 across all t tests).

To examine the unique and interactive contributions of emo-tional flooding and affect appraisal bias to negative parentingpractices, a series of linear regression models were tested in whicheach parenting domain was examined in turn. The dependentvariable (DV) in each model was the parenting domain (Model 1:over reactive discipline; Model 2: lax discipline: Model 3 hostilediscipline). The same set of independent variables (IV) was testedin each. These were entered in three blocks, the first of whichcomprised child age, single parent status, and disruptive behaviorproblem severity (ECBI Intensity score). Block two comprised theparent cognition variables of emotional flooding, and affect ap-

Table 1Descriptive Statistics and Correlations Among Key Child and Parent Variables

Variable 1 2 3 4 5 6 7 8 9

1. Child age 12. Observed child anger .03 13. Affect appraisal bias �.03 .13 14. Emotional flooding .22� .01 .23� 15. Overreactive discipline .18 .07 .04 .13 16. Lax discipline �.01 �.03 �.01 .20 .04 17. Hostile discipline .08 .09 .24� .39�� .35�� .29�� 18. ECBI intensity scale .42�� .07 .23� .43�� .30�� .06 .24� 19. ECBI problem scale .46�� .12 .25� .48�� .29� .06 .20 .75�� 1M 31.53 3% 2.81 22.53 3.64 3.47 1.97 128.84 13.32SD 9.18 5.22 .73 6.27 1.11 1.12 1.00 36.99 8.06� — — — .86 .63 .67 .68 .94 .88

Note. ECBI � Eyberg Child Behavior Inventory.� p � .05. �� p � .01.

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

16 MENCE ET AL.

praisal bias, whereas bock three comprised the two-way interactionterm for affect appraisal bias � emotional flooding.

As recommended when testing interaction terms in regression(Aiken & West, 1991), all predictor variables were first centered.The Variance Inflation Factor (VIF; Velleman, 1981) was alsocalculated for each model, to check that they were not affected bymulticollinearity. The maximum value among all regressions con-ducted (including those described later) was 4.25; models aredeemed uninterpretable when VIF values reach 10 (Belseley, Kuh,& Welsch, 1980). Significant interaction terms were probed posthoc using the established method of simple slope analysis outlinedby Aiken and West (1991), which is recommended for interpretingmoderational effects in child and family research (Holmbeck,2002). Using this method, conditional moderator variables, corre-sponding to � 1 SD from the centered value for each participant,were computed to test the significance of the respective IV athigh/low levels of the moderator variable.

Overreactive discipline (Model 1) was found to be significantlypredicted by severity of disruptive behavior problems (� � .24,SE � .01, p � .05), but no other variables. Because no significanteffects were seen in blocks 2 or 3, block 1 of the model only wasinterpreted (see Table 2). Hostile discipline (Model 2) was foundto be predicted by emotional flooding in block 2 (� � .28, SE �

.05, p � .05), as well as the interaction term for affect appraisalbias � emotional flooding in block 3 (� � .20, SE � .14, p � .05).Regression statistics for the full model are reported in Table 2. Posthoc analysis of this interaction effect revealed that affect appraisalbias was significantly associated with hostile discipline amongparents who reported high levels of emotional flooding (� � .40,SE � .67, p � .05), but not among parents who reported low levelsof emotional flooding (� � �.03, SE � .57, ns). This interactionis shown in Figure 1. Lax discipline (Model 3) was not found to beassociated with any of the predictor variables tested. Regressionstatistics for block 1 of this model are reported in Table 2.

Discussion

This study examined the relationship between negative parent-ing practices and dysfunction in parents’ processing of child affectin families of toddlers referred for the treatment of disruptivebehavior problems. As hypothesized, the more that parents in thesefamilies were prone to emotional flooding, the more likely theywere to misclassify child affect as anger. As such, the biasedappraisal of child affect cues was most pronounced among thoseparents who found their child’s negative affect to be unpredictableand overwhelming, and who experienced overarousal and disrup-

Table 2Unique Predictors of Overreactive, Hostile, and Lax Discipline

Block Predictor variables

Overreactivediscipline (Model 1)

Hostile discipline(Model 2)

Lax discipline(Model 3)

SE � SE � SE �

1 Child age 0.07 0.09 0.03 �0.02 0.07 �0.05Single parent status 2.07 0.03 0.97 0.11 1.86 0.20Socioeconomic index 0.33 0.08 0.16 0.08 0.29 0.00Eyberg Child Behavior Inventory (Intensity) 0.01 0.25� 0.10 0.03 0.01 0.03

2 Emotional flooding — — 0.05 0.31� — —Affect appraisal bias — — 0.45 0.18 — —

3 Affect appraisal bias � Emotional flooding — — 0.07 0.20� — —

� p � .05. �� p � .01.

Affect Appraisal Bias

3.00

4.00

5.00

6.00

7.00

8.00

Lo AngerBias Med AngerBias Hi AngerBias

Lo Flooding

Med Flooding

Hi Flooding

Low(-1 SD)

Moderate High(+1 SD)

High Flooding (+1 SD)

Moderate Flooding

Low Flooding (-1 SD)

Hos

tile

Dis

cipl

ine

Figure 1. Interaction between emotional flooding and affect appraisal bias in predicting hostile discipline.

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

17EMOTIONAL FLOODING AND HOSTILE DISCIPLINE

tion to higher-order cognition (e.g., problem-solving capacities)when confronted with this affect. This association was significanteven after controlling for individual differences in severity ofchildren’s disruptive behavior problems.

Both hostile and overreactive discipline were positively associ-ated with the severity of disruptive behavior problems in oursample, however only hostile discipline was associated with par-ents’ biased appraisal of child affect and emotional flooding. It isnoteworthy that we did not find an association between affectiveappraisal bias and overreactive parenting, when such an associa-tion has been reported in previous studies (Lorber & O’Leary,2005; Lorber et al., 2003). It is possible that this discrepancy maystem from sample differences, with these previous studies con-ducted with nonclinical samples. Additionally, it is possible thatthe observational measures of overreactivity used in these previousstudies may have tapped parent behaviors that overlap more withthe self-report measure of hostile discipline in the current study,than the self-report measure of overreactive discipline. Unlikethese previous studies, which examined overreactivity in isolationfrom hostility, the current study measured both dimensions usingself-report items based on factor analytic evidence of specificityregarding features of overreactive versus hostile discipline(Rhoades & O’Leary, 2007). Future research using designs thatincorporate both clinic-referred and community samples is neededto better understand this pattern of associations, and the conditionsunder which they may vary. Our current data nonetheless suggestthat in families of young children with clinic-referred disruptivebehavior problems, the biased processing of child affect may besomewhat uniquely related to a particularly hostile form of disci-pline involving the use of physical or verbal force.

In addition to finding that affect appraisal bias and emotionalflooding were each characteristic of parents who engaged in highlevels of hostile discipline, our results suggest that processesinvolving the interplay between these two variables contribute torisk for hostile discipline. Specifically, we found that hostilediscipline was predicted by the interaction between affect appraisalbias and emotional flooding. High levels of flooding were associ-ated with a particularly strong risk for hostile discipline amongparents who also exhibited an affect appraisal bias. Although anaffect appraisal bias was associated with hostile discipline amongparents who were prone to emotional flooding, no such associationwas seen with this bias among parents who were free from flood-ing. It is interesting to consider this interaction in relation to themixed findings that have emerged from research into emotionrecognition biases among parents at risk for physical abuse (Aslaet al., 2011). Our data suggest that one explanation for those mixedfindings may be that previous research samples have varied interms of the distribution of emotional flooding therein.

Gottman’s (1991, 1993) conceptualization of emotional flood-ing has been a noteworthy influence on models of family mecha-nisms in the development and maintenance of disruptive behaviorproblems (e.g., Snyder et al., 1994), however it has been the focusof relatively little child and family research. Previous research hasshown that parent variables closely related to experiential compo-nents of emotional flooding, such as physiological arousal elicitedby negative child affect, are related to negative parenting practicesin community samples of families with young children (e.g.,Lorber & O’Leary, 2005; Lorber & Slep, 2005). Likewise, longi-tudinal evidence from research into parenting and children’s emo-

tion socialization in the family indicates that the tendency to reactwith distress to children’s negative affect is a risk factor for laterexternalizing problems (Eisenberg et al., 1999). Research in thisarea has further shown that these distress reactions are highlystable over time, and associated also with punitive responses tochildren’s expressions of negative affect (Eisenberg et al., 1999;Fabes et al., 2002). However, only two previous studies of parent-ing have operationalized emotional flooding specifically, both ofwhich were conducted with nonclinical samples (Lorber et al.,2011; Slep & O’Leary, 2007). The findings of the current studyprovide preliminary evidence that emotional flooding explainsindividual differences in quality of parenting among children re-ferred for treatment of disruptive behavior problems in the earlychildhood years.

It is noteworthy that the current study was conducted in collab-oration with a tertiary-level health service—that is, one in whichclients require a referral from another health professional such asa pediatrician or general practitioner, to access treatment. Giventhe demands associated with observational research with families,research into parenting processes in families of children withdisruptive behavior problems has often been conducted inuniversity-based clinics, with self-referred families who arethought to be higher functioning that those typically seen in “realworld” community health settings. As such, a particular strength ofthe current study is that the findings can be considered generaliz-able to such real world settings.

The way in which observational coding was used in the currentstudy might raise the question of whether the parents who reportedhigher levels of child affect than the independent observers mayindeed have been more sensitive to the meaning of their children’saffective cues than those observers. However, this seems unlikely,if we are to assume that parents who are more sensitive to themeaning of their children’s affect would also respond to theirchildren’s affect and behavior more sensitively. In contrast, wefound that the more that parents’ appraisals of their children’saffect deviated from those of independent observers, the morelikely they were to engage in hostile discipline. Conversely,greater agreement between parents and observers was associatedwith lower levels of hostility.

The results of the current study should be interpreted in light ofa number of limitations. First, the design of this study was cross-sectional, thereby preventing examination of temporal associationsbetween parenting practices and parents’ processing of child af-fect. It is possible, for example, that although flooding was notassociated with concurrent levels of overreactive or lax discipline,parents who are prone to flooding may become increasingly over-reactive or lax over time. Likewise, the cross-sectional designprecludes conclusive interpretations regarding directions of causalinfluence. Second, the parent variables examined here were in-dexed in the majority of cases using data collected from mothers.Evidence from previous research into emotional flooding in non-clinical families suggests that the associations between emotionalflooding and other parent and child variables may vary somewhatbetween mothers and fathers (Slep & O’Learly, 2007). As such,future research is needed to test whether our findings generalize tothe parenting of fathers of young children referred for disruptivebehavior problems. Finally, although our measure of affect ap-praisal bias can be considered free from reporter bias because ofthe multimethod strategy involving independent coding of child

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

18 MENCE ET AL.

affect, our measure of emotional flooding was self-report andtherefore shared method variance with our measure of parentingpractices. This is a limitation that is shared by previous researchinto parenting and emotional flooding (Slep & O’Leary, 2007),and would be of value to overcome in future research by devel-oping methods to operationalize the unique components of emo-tional flooding using physiological or experimental paradigms.

Given the limited research that has investigated parental flood-ing to date, it is not surprising that little has been published onmeasurement issues related to the assessment of the construct.Emotional flooding and hostile discipline showed distinct associ-ations with other conceptually related variables in the current dataset, supporting the assumption that the measure of emotionalflooding used in this study was not simply measuring a feature ofhostile discipline.1 However, at the same time, future measurementresearch into emotional flooding is needed if researchers are tofurther investigate the interplay between flooding and relatedparenting processes.

It is worth noting that a relatively low base rate for child angerwas observed in the sample as a whole. Given that this investiga-tion was concerned with negative bias in parents’ appraisal of childaffect (i.e., the tendency to misclassify nonangry affect as expres-sions of anger), this low base rate was not incompatible with theaims of the study. However, it is indeed possible that the results ofthe study may have differed in a sample characterized by morefrequent expressions of anger. It would therefore be of interest toexamine how the processes identified in the current study mayvary across samples characterized by distinct rates of anger ex-pression in future research.

It is critical that the factors underlying hostile and aggressiveparenting are understood by clinicians working in early childhoodsettings, with national incidence data showing that young childrenare at the greatest risk of physical maltreatment by parents (Sedlak& Broadhurst, 1996). Hostile discipline is a particularly noxiousaspect of the coercive parent–child exchanges that have beenshown to amplify and maintain chronic trajectories of antisocialbehavior across development, and in many cases these coercivecycles first emerge in the early childhood years (Patterson, 2002).Importantly also, it is during the toddler years that children’sdisruptive behavior most commonly features intense displays ofnegative affect (Cole, Teti, & Zahn-Waxler, 2003). Our findingsadd to growing evidence that dysfunction in parents’ processing ofchild affect contributes significantly to hostile discipline, and showfor the first time that emotional flooding may be particularlyproximal to such parenting in the families of toddlers with disrup-tive behavior problems.

The model supported by our data suggests that a parent who isbiased toward “seeing” more anger in a child than is present, maynonetheless be able to maintain healthy parenting practices so longas this misappraisal does not cascade into the physiological arousaland subsequent cognitive “shutdown” that characterizes emotionalflooding. As such, a key clinical implication of our findings is thattherapeutic efforts to address parents’ processing of child affectcues in their children may be most likely to reduce harsh parentingif these efforts target emotional flooding as a proximal mechanism,rather than attempting to simply improve the accuracy with whichparents are able to recognize their children’s affective cues. Inclinical work with distressed couples—the population in whichrisk-processes related to emotional flooding were first conceptu-

alized—it has been recommended that partners take “time out”from each other when experiencing emotion flooding, to allow theassociated arousal to dissipate before engaging in further interac-tions (Gottman, 1994). In much the same way, providing parentswith skills for the effective implementation of time-out with chil-dren (e.g., briefly removing a child from social reinforcement) maybe particularly beneficial in preventing the escalation of hostilityamong parents prone to flooding. Where possible, the utilization ofsocial support that allows a flooded parent to have time-out froma child is also likely to be of considerable benefit.

As previously proposed by Slep and O’Leary (2007), it may alsobe of clinical value to target emotional flooding more directly byteaching parents distress-tolerance skills to cope with the diffusearousal that characterizes the reaction. Various parent trainingprograms for disruptive behavior problems have been developed inwhich parents are first introduced to child-focused modules tar-geting behavior management strategies, before moving on toparent-focused modules targeting issues such as marital discordand parental mood and anger problems (e.g., Dadds & Hawes,2006). It would seem feasible to incorporate skills training forcoping with emotional flooding within such a structure, withparents supported in their implementation of such skills once theyhave achieved sufficient mastery of core skills in reinforcingdesirable child behavior and setting limits on oppositional andaggressive behavior.

The current study found that among parents of toddlers withdisruptive behavior problems, those who are prone to emotionalflooding in response to negative child affect are at risk for aparticularly hostile pattern of discipline. Furthermore, results sug-gest that a negative bias in parents’ appraisal of the affective cuesof family members, which has previously been found to charac-terize parents of children with disruptive behavior problems (e.g.,Sanders et al., 1992) as well as parents at risk of committing abuse(e.g., Asla et al., 2011), is only likely to contribute to hostilediscipline when accompanied by emotional flooding. Our findingthat emotional flooding is particularly proximal to hostile disci-pline in the families of toddlers with clinic-referred behaviorproblems is consistent with evidence previously reported for non-clinical families (Slep & O’Leary, 2007), and supports the poten-tial value of targeting parents’ capacities to cope with emotionalflooding in clinical interventions for disruptive behavior problems.

References

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing andinterpreting interactions. Newbury Park: Sage.

Asla, N., de Paúl, J., & Pérez-Albéniz, A. (2011). Emotion recognition infathers and mothers at high risk for child physical abuse. Child Abuse &Neglect, 35, 712–721. doi:10.1016/j.chiabu.2011.05.010

Arnold, D. S., O’Leary, S. G., Wolff, L. S., & Acker, M. M. (1993). Theparenting scale: A measure of dysfunctional parenting in disciplinesituations. Psychological Assessment, 5, 137–144. doi:10.1037/1040-3590.5.2.137

Australian Bureau of Statistics. (1997). Australian standard classificationof occupations (ASCO) (2nd ed.). Canberra, Australia: Australian Bu-reau of Statistics.

Australian Bureau of Statistics. (2007). Census of population and housing:2006 census tables (No. 2068.0). Canberra, Australia: Australian Bureauof Statistics.

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

19EMOTIONAL FLOODING AND HOSTILE DISCIPLINE

Belseley, D. A., Kuh, E., & Welsch, R. E. (1980). Regression diagnostics:Identifying influential data and sources of collinearity. New York:Wiley.

Boggs, S. R., Eyberg, S. M., & Reynolds, L. (1990). Concurrent validity ofthe Eyberg Child Behavior Inventory. Journal of Clinical Child Psy-chology, 19, 75–78. doi:10.1207/s15374424jccp1901_9

Burns, G. L., & Patterson, D. R. (1990). Conduct problem behaviors in astratified random sample of children and adolescents: New standardiza-tion data on the Eyberg Child Behavior InventoryPsychological Assess-ment: A Journal of Consulting and Clinical Psychology, 2, 391–397.doi:10.1037/1040-3590.2.4.391

Coan, J. A., & Gottman, J. M. (2007). The Specific Affect (SPAFF) codingsystem. In J. A. Coan & J. J. B. Allen (Eds.), Handbook of emotionelicitation and assessment (pp. 106–123). New York, NY: OxfordUniversity Press.

Cole, P. M., Teti, L. O., & Zahn-Waxler, C. (2003). Mutual emotionregulation and the stability of conduct problems between preschool andearly school age. Development and Psychopathology, 15, 1–18. doi:10.1017.S0954579403000014

Dadds, M. R., & Hawes, D. J. (2006). Integrated family intervention forchild conduct problems. Brisbane, Australia: Australian Academic Press.

Dishion, T. J., & Patterson, G. R. (2006). The development and ecology ofantisocial behavior in children and adolescents. In D. Cicchetti & D. J.Cohen (Eds.), Developmental psychopathology (pp. 503–541). NewYork, NY: Wiley.

Dix, T. (1991). The affective organization of parenting: Adaptive andmaladaptive processes. Psychological Bulletin, 110, 3–25. doi:10.1037/0033-2909.110.1.3

Duppong-Hurley, K., Huscroft-D’Angelo, J., Tout, A., Griffith, A., &Epstein, M. (2013). Assessing parenting skills and attitudes: A review ofthe psychometrics of parenting measures. Journal of Child and FamilyStudies. Advance online publication. doi:10.1007/s.10826-013-9733-2

Eisenberg, N., Fabes, R. A., Shepard, S. A., Guthrie, I. K., & Murphy,B. C., & Reiser, M. (1999). Parental reactions to children’s negativeemotions: Longitudinal relations to quality of children’s social function-ing. Child Development, 70, 513–534. doi:10.1111/1467-8624.00037

Ekman, P. (1984). Expression and the nature of emotion. In K. Scherer &P. Ekman (Eds.), Approaches to emotion (pp. 319–343). Hillsdale, NJ:Erlbaum.

Eyberg, S. M., & Pincus, D. (1999). Eyberg Child Behavior Inventory andSutter-Eyberg Student Behavior Inventory-revised: Professional man-ual. Odessa, FL: Psychological Assessment Resources.

Fabes, R. A., Poulin, R. E., Eisenberg, N., & Madden-Derdich, D. A.(2002). The Coping with Children’s Negative Emotions Scale (CCNES):Psychometric properties and relations with children’s emotional compe-tence. Marriage & Family Review, 34, 285–310. doi:10.1300/J002v34n03_05

Gottman, J. M. (1991). Chaos and regulated change in families: A meta-phor for the study of transitions. In P. A. Cowan & M. Hetherington(Eds.), Family transitions (pp. 247–272). Hillsdale, NJ: Erlbaum.

Gottman, J. M. (1993). A theory of marital dissolution and stability.Journal of Family Psychology, 7, 57–75. doi:10.1037/0893-3200.7.1.57

Gottman, J. M. (1994). Why marriages succeed or fail. New York, NY:Simon & Schuster.

Hawes, D. J., & Dadds, M. R. (2005). Oppositional and conduct problems.In J. Hudson & R. Rapee (Eds.), Current thinking on psychopathologyand the family (pp. 73–91). New York, NY: Elsevier. doi:10.1016/B978-008044449-9/50006-X

Hawes, D. J., & Dadds, M. R. (in press). Parent and family assessmentstrategies. In B. D. McLeod, A. Jensen-Doss, & T. Ollendick (Eds.),Handbook of child and adolescent diagnostic and behavioral assess-ment. New York, NY: Guilford Press.

Hawes, D. J., Dadds, M. R., & Pasalich, D. (in press). Observational coding

strategies. In J. Comer & P. Kendall (Eds.), The Oxford handbook ofresearch strategies for clinical psychology. New York, NY: OxfordUniversity Press.

Holmbeck, G. N. (2002). Post-hoc probing of significant moderational andmediational effects in studies of pediatric populations. Journal of Pedi-atric Psychology, 27, 87–96. doi:10.1093/jpepsy/27.1.87

Johnston, C., & Ohan, J. L. (2005). The importance of parental attributionsin families of children with attention-deficit/hyperactivity and disruptivebehavior disorders. Clinical Child and Family Psychology Review, 8,167–182. doi:10.1007/s10567-005-6663-6

Lorber, M. F., & O’Leary, S. G. (2005). Mediated paths to over-reactivediscipline: Mothers’ experienced emotion, appraisals, and physiologicalresponses. Journal of Consulting and Clinical Psychology, 73, 972–981.doi:10.1037/0022-006X.73.5.972

Lorber, M. F., O’Leary, S. G., & Kendziora, K. T. (2003). Mothers’over-reactive discipline and their encoding and appraisals of toddlerbehavior. Journal of Abnormal Child Psychology, 31, 485–494. doi:10.1023/A:1025496914522

Lorber, M. F., O’Leary, S. G., & Slep, A. M. S. (2011). An initialevaluation of the role of emotion and impulsivity in explaining racial/ethnic differences in corporal punishment. Developmental Psychology,47, 1744–1749. doi:10.1037/a0025344

Lorber, M. F., & Slep, A. M. S. (2005). Mothers’ emotion dynamics andtheir relations with harsh and lax discipline: Microsocial time seriesanalyses. Journal of Clinical Child and Adolescent Psychology, 34,559–568. doi:10.1207/s15374424jccp3403_11

Milner, J. S. (1993). Social information processing and physical childabuse. Clinical Psychology Review, 13, 275–294. doi:10.1016/0272-7358(93)90024-G

Moffitt, T. E., & Caspi, A. (2007). Evidence from behavioral genetics forenvironmental contributions to antisocial conduct. In J. Grusec & P.Hastings (Eds.), Handbook of socialization (pp. 96–123). New York,NY: Guilford Press.

Patterson, G. R. (1982). Coercive family process. Eugene, OR: Castalia.Patterson, G. R. (2002). The early development of coercive family process.

In J. B. Reid, G. R. Patterson, & J. J. Snyder (Eds.), Antisocial behaviorin children and adolescents: A developmental analysis and model forintervention (pp. 25–44). Washington, DC: American PsychologicalAssociation. doi:10.1037/10468-002

Rhoades, K. A., & O’Leary, S. G. (2007). Factor structure and validity ofthe Parenting Scale. Journal of Clinical Child and Adolescent Psychol-ogy, 36, 137–146. doi:10.1080/15374410701274157

Rich, B. A., & Eyberg, S. M. (2001). Accuracy of assessment: Thediscriminative and predictive power of the Eyberg Child BehaviorInventory. Ambulatory Child Health, 7, 249–257. doi:10.1046/j.1467-0658.2001.00141.x

Roberts, W., & Strayer, J. (1987). Parents’ responses to the emotionaldistress of their children: Relations with children’s competence. Devel-opmental Psychology, 23, 415–422. doi:10.1037/0012-1649.23.3.415

Sanders, M. R., Dadds, M. R., Johnson, B. M., & Cash, R. (1992).Childhood depression and conduct disorder: I. Behavioral, affective, andcognitive aspects of family problem-solving interactions. Journal ofAbnormal Psychology, 101, 495–504. doi:10.1037/0021-843X.101.3.495

Sedlak, A. J., & Broadhurst, D. D. (1996). Third national incidence studyof child abuse and neglect. Washington, DC: National Center on ChildAbuse and Neglect.

Siever, L. J. (2008). Neurobiology of aggression and violence. The Amer-ican Journal of Psychiatry, 165, 429–442. doi:10.1176/appi.ajp.2008.07111774

Slep, A. M. S., & Heyman, R. E. (1998). Parental flooding scale. Unpub-lished manuscript, Stony Brook University, State University of NewYork.

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

20 MENCE ET AL.

Slep, A. M. S., & O’Leary, S. G. (2007). Multivariate models of mothers’and fathers’ aggression toward their children. Journal of Consulting andClinical Psychology, 75, 739–751. doi:10.1037/0022-006X.75.5.739

Smith, A. M., & O’Leary, S. G. (1995). Attributions and arousal aspredictors of maternal discipline. Cognitive Therapy and Research, 19,459–471. doi:10.1007/BF02230412

Snarr, J. D., Strassberg, Z., & Slep, A. M. S. (2003). Making faces: Testingthe relation between child behavior problems and mothers’ interpreta-tions of child emotion expressions. Journal of Abnormal Child Psychol-ogy, 31, 371–380. doi:10.1023/A:1023887401140

Snyder, J., Edwards, P., McGraw, K., Kilgore, K., & Holton, A. (1994).Escalation and reinforcement in mother-child conflict: Social processesassociated with the development of physical aggression. Developmentand Psychopathology, 6, 305–321. doi:10.1017/S0954579400004600

Snyder, J., Schrepferman, L., McEachern, A., & DeLeeuw, J. (2010). Thecontribution of child anger and fear, and parental discipline to earlyantisocial behavior: An integrative model. In M. Portegal, G. Stemmler,& C. Spielberger (Eds.), Handbook of anger: Biological, psychological,

and social processes. New York, NY: Springer. doi:10.1007/978-0-387-89676-2_29

Velleman, P. F. (1981). Efficient computing of regression diagnostics.American Statistician, 35, 234–242. doi:10.2307/2683296

Vondra, J., & Belsky, J. (1993). Developmental origins of parenting:Personality and relationship factors. In T. Luster & L. Okagaki (Eds.),Parenting: An ecological perspective (pp. 1–33). Hillsdale, NJ: Erl-baum.

Wahler, R. G., & Dumas, J. E. (1989). Attentional problems in dysfunc-tional mother-child interactions: An interbehavioral model. Psycholog-ical Bulletin, 105, 116–130. doi:10.1037/0033-2909.105.1.116

Received February 11, 2013Revision received October 15, 2013

Accepted October 22, 2013 �

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

21EMOTIONAL FLOODING AND HOSTILE DISCIPLINE