Method
Search Strategy and Inclusion Criteria
Data Extraction
Results
Methodological Aspects of Included Studies
Study | Aim | Observational coding system | Observation task | Coder information | Sample | Significant group differences in parenting behavior of maltreating versus nonmaltreating parents | Effect size (Cohen’s d) |
---|---|---|---|---|---|---|---|
Cicchetti et al. (2006) US | Detect differences in the efficacy of three different preventive interventions in mother–child dyads that do versus do not engage in maltreatment | MBQS | 3-hr home observation | Two trained coders Intraclass correlations between pairs of observers 0.71 to 0.72 | N = 189 mother–child dyads, n = 137 dyads with substantiated history of abuse Child age in months M = 13.31 (SD = 0.81); Est. range: 10.88–15.74 months 74.1% ethnic minority | Significantly lower sensitivity scores for mothers engaging in maltreatment (M = 0.27, SD = 0.36) than mothers not engaging in maltreatment (M = 0.48, SD = 0.29). | d = − 0.39 |
Cipriano-Essel et al., (2013) US | Detect differences in parent and child behavior by comparing mother–child dyads who do versus do not engage in maltreatment | SASB | 3–5 min moderately challenging joint teaching task | Trained coders (75 h), blind to maltreatment status Weighted kappas 0.53 to 1.0 (M = 0.74) | N = 118 mother–child dyads, n = 64 dyads with substantiated history of abuse Child age in years: M = 3.7 (SD = 0.74); Range: 3–5 years 79.8% Caucasian, 2.5% African American, 16.2% multiracial | Mothers engaging in maltreatment (M = 0.31, SD = 0.21) show significantly more strict/hostile control than mothers not engaging in maltreatment (M = 0.23, SD = 0.21). | d = 0.38 |
Fagan & Dore (1993) US | Examining differences in mother–child play between mother–child dyads who do versus do not engage in neglect | P/CIS | 30-min free play session at home | Trained research assistants and experienced child welfare professionals, blind to the maltreatment status and purpose of the study | N = 27 mother–child dyads; n = 17 neglected children Child age in months M = 27.5 (SD and range not reported) 62% African American, 38% Caucasian | Significantly lower maternal positive responsiveness for mothers engaging in neglect (M = 16.00) than for mothers not engaging in neglect (M = 20.54). | n. a. (SD not reported) |
Significantly lower maternal development appropriateness for mothers engaging in neglect (M = 37.29) than for mothers not engaging in neglect (M = 53.89). | n. a. (SD not reported) | ||||||
Haskett et al. (2008) US | Examining protective and vulnerability factors for physically abused children | Qualitative ratings of parent–child interactions | Three different tasks, 30 min | Trained coders, uninformed of maltreatment status κ = 0.80 | N = 153 parent–child dyads; n = 78 dyads with substantiated history of abuse Child age in years: M = 7.2 (SD = 1.4); Range: 3.0–11.4 69% African American, 31% Caucasian | Significantly lower scores for parental sensitivity in parents engaging in maltreatment (M = 3.51, SD = 1.3) than in parents not engaging in maltreatment (M = 3.98, SD = 1.3). | d = − 0.36 |
Hurlburt et al. (2013) US | Examining efficacy of an intervention program in families with self-reported history of child maltreatment | DPICS-II & CII | 30-min home observation | Highly experienced or extensively trained coders Intraclass correlations 0.70 to 0.87 | N = 481 parent–child dyads; n = 100 dyads with self-reported history of abuse Child age in years: M = 4.7 (SD = 0.36); Est. range: 3.62–5.78 years 60.3% Caucasian, 8.4% African American, 4.6% Hispanic/Latino, 1.0% Asian American, 1.5% Native American, 14.2% multiracial | DPICS-II: Significantly more total critical statements in parents engaging in maltreatment (M = 25.48, SD = 19.30) than in parents not engaging in maltreatment (M = 19.07, SD = 17.36). | d = 0.36 |
CII: Significantly less nurturing/supportive behavior in parents engaging in maltreatment (M = 2.29, SD = 0.53) versus parents not engaging in maltreatment (M = 2.47, SD = 0.45). | d = − 0.39 | ||||||
CII: Significantly more harsh/critical behavior in parents engaging in maltreatment (M = 1.60, SD = 0.58) versus parents not engaging in maltreatment (M = 1.42, SD = 0.46). | d = 0.37 | ||||||
CII: Significantly less discipline competence in parents engaging in maltreatment (M = 2.11, SD = 0.49) versus parents not engaging in maltreatment (M = 2.22, SD = 0.41). | d = − 0.26 | ||||||
Lau et al. (2006) US | Examining relationships between abusive versus nonabusive parents’ reports of child behavior and observed parent–child interactions | UCLA Parent–Child Coding System | Three different interaction tasks, 13 min | Trained undergraduate research assistants achieving reliability of at least 70% | N = 205 parent–child dyads; n = 58 dyads with substantiated history of abuse Child age in years: M = 11.5 (SD = 2.65); Est. range: 3.55–19.45 years Sample with substantiated history of abuse: 36.2% Caucasian, 27.6% African American, 12.1% Hispanic American, 24.1% other Sample without substantiated history of abuse: 45.6% Caucasian, 15.6% African American, 14.3% Hispanic American, 2.7& Asian American, 21.8% other | After controlling for family income, child age and sex, and parental psychopathology, parents engaging in abuse showed more emotionally controlling behaviors (e.g., criticism, guilt induction, and intrusiveness). | n. a. (M and SD not reported) |
Parents engaging in abuse showed less supportive behavior (e.g., praise, encouragement, and display of affection). | n. a. (M and SD not reported) | ||||||
Ostler (2010) US | Determine risk of reoccurrence of maltreatment; examination of reliability and validity of CARE-Index | CARE-Index | 3-min free play session in clinical setting | Independent and trained coders κ = 0.90 for 30 protocols | N = 44 mother–child dyads confirmed for maltreatment; n = 27 control group dyads Sample of mothers with chronic and severe mental illness Child mean age not reported; age range: 8–48 months 55% African American, 36% Caucasian, 7% Hispanic, 2% Asian American | Scores on the CARE-Index were associated with the comprehensive measure of risk. Sensitivity scores were significantly lower for mothers at high risk (M = 4.7, SD = 2.6) than for mothers at low risk (M = 8.5, SD = 1.5). | d = − 1.55 |
Scores for maternal sensitivity were significantly lower for mothers at high risk (M = 4.7, SD = 2.6) than for mothers at moderate risk (M = 6.1, SD = 2.1). | d = − 0.57 | ||||||
Robinson et al. (2009) US | Detecting relations between emotion regulation, parenting, and psychopathology in maltreated children | Parent–Child Interaction Procedure | Seven different tasks with different levels of difficulty, 30–45 min | Trained coders Interrater reliability 71–84% | N = 123 parent–child dyads; n = 66 dyads with substantiated history of abuse Child age in months: M = 32.62 (SD = 9.71); Range: 12–47 months 69% African American, 31% Caucasian, Hispanic, biracial or other | Significantly lower parental positive affect intensity for parents engaging in maltreatment (M = 3.77, SD = 0.92) versus parents not engaging in maltreatment (M = 4.63, SD = 0.56). | d = − 1.11 |
Significantly higher parental anger intensity for parents engaging in maltreatment (M = 1.66, SD = 0.81) versus parents not engaging in maltreatment (M = 1.16, SD = 0.27). | d = 0.81 | ||||||
Sabourin Ward & Haskett (2008) US | Exploration and validation of clusters of children who suffered from physical abuse | Qualitative ratings of parent–child interactions | Three different tasks, 30 min | Trained, uninformed of maltreatment status κ = 0.80 for reliability between raters | N = 175 parent–child dyads; n = 98 dyads with substantiated history of abuse Child age in years: M = 7.33 (SD = 1.54); Range: 5–10 years 70% African American, 27% European American, 3% Hispanic or biracial | Parents engaging in maltreatment (M = 6.93, SD = 2.6) showed significantly less sensitivity than parents not engaging in maltreatment (M = 8.09, SD = 2.6). | d = − 0.45 |
Skowron et al. (2011) US | Investigate mother and child physiological regulation (RSA) and their associations to observed parent–child interaction | SASB | Mother–child joint task | Trained coders (75 hrs. of training), blind to family status κ from 0.64 to 0.84 (M = 0.74) for reliability between raters | N = 76 children and their mothers; n = 17 suffered from physical abuse; n = 26 suffered from physical neglect Child age in years: M = 3.7 (SD = 0.73); Range: 3–5 years 77.3% Caucasian, 18.7% multiracial, 4.0% African American | Mothers engaging in physical abuse (M = 0.24, SD = 0.15) were significantly more likely to react with strict control than mothers who did not engage in maltreatment (M = 0.13, SD = 0.13). | d = 0.80 |
Mothers engaging in neglect (M = 0.23, SD = 0.16) were significantly more likely to react with strict control than mothers who did not engage in maltreatment (M = 0.13, SD = 0.13). | d = 0.70 | ||||||
Mothers engaging in physical abuse (M = 0.12, SD = 0.09) were significantly less likely to affirm autonomy to their children than mothers who did not engage in maltreatment (M = 0.22, SD = 0.13). | d = − 0.85 | ||||||
Skowron et al. (2013) US | Examine the relationship between concurrent and lagged physiological regulation (RSA) and parenting behavior of maltreating and nonmaltreating mothers | SASB | Mother–child teaching task | Trained coders (75 h of training), blind to maltreatment status κ from 0.53 to 1.0 (M = 0.74) for reliability between raters | N = 141 mothers and their children; n = 62 maltreating/neglecting mothers Child age in years: M = 3.7 (SD = 0.72); Range: 3–5 years 94.2% Caucasian | Mothers engaging in abuse (M = 32.98) and mothers engaging in neglect (M = 35.28) showed less positive parenting than mothers who did not engage in maltreatment (M = 42.52). | n. a. (SD not reported) |
Mothers engaging in abuse (M = 0.34) and mothers engaging in neglect (M = 0.32) mothers showed more strict/hostile control than mothers who did not engage in maltreatment (M = 0.20). | n. a. (SD not reported) |
Study | Aim | Observational coding system | Observation task | Coder information | Sample | Classification of parents as maltreating versus nonmaltreating | Sensitivity and specificity of classification |
---|---|---|---|---|---|---|---|
Brassard et al. (1993) US | Evaluate the factor structure and validity of the Psychological Maltreatment Rating Scales (PMRS) | PMRS | 15-min teaching task | Trained research assistants | N = 49 mother–child dyads; n = 25 children who suffered from maltreatment Child age in years: M = 6.67 (SD not reported); Range: 4–8.92 years 46% Caucasian, 41% African American, 13% Hispanic | Coding system correctly classified 81.63% of the mothers as engaging in psychological maltreatment versus not engaging in psychological maltreatment | Sensitivity = 0.92 Specificity = 0.71 |
Cerezo et al. (1996) Spain | Detecting the differentiating aspects of abusive and nonabusive mothers’ behavior during parent–child interaction | SOC-III | Several 1-hr unstructured interactions in home | Trained undergraduate psychology students (average 20 hr of training) | N = 47 mother–child dyads; n = 23 children who suffered from maltreatment Sample with maltreatment history: Child age in years: M = 8.04 (SD = 2.78); Est. range: 0–16.38 years Sample without maltreatment history: Child mean age in years: M = 8.0 (SD = 2.16); Est. range: 1.52–14.48 years Ethnicity: n. a. | Maternal behavior scale correctly identified 70.21% of the mothers as engaging in maltreatment versus not engaging in maltreatment | Sensitivity = 0.79 Specificity = 0.61 |
Child behavior scale correctly identified 74.47% of the mothers as engaging in maltreatment versus not engaging in maltreatment | Sensitivity = 0.88 Specificity = 0.61 | ||||||
Mother–child interactional behavior scale correctly identified 78.72% of the mothers as engaging in maltreatment versus not engaging in maltreatment | Sensitivity = 0.88 Specificity = 0.61 | ||||||
The whole model correctly identified 82.98% of the mothers as engaging in maltreatment versus not engaging in maltreatment | Sensitivity = 0.88 Specificity = 0.78 |
Sample Characteristics of Individual Studies
Differences in Behavior of Parents Who Did Versus Did Not Engage in Child Maltreatment
Classification of Parents as Engaging Versus Not Engaging in Maltreatment
Characteristics of Observational Coding Systems
Observational coding system | Author (year) | Age range | Measured construct | Dimensions | Subscales | Observation task | Reliability of coding system as reported in included studies |
---|---|---|---|---|---|---|---|
CARE-Index | 0–15 months (15–30 months; toddler version) | Maternal sensitivity Child’s interactional behavior | Sensitivity Hostile control Unresponsiveness Cooperative Compulsive Compliance Difficult (threatening; toddler version) Passive (disarming; toddler version) | Facial expression Vocal expression Position and body contact Expression of affection Pacing of turns Developmental appropriateness of activity Facial expression Vocal expression Position and body contact Expression of affection Pacing of turns Developmental appropriateness of activity | 3-min free play session in a clinical setting | Interrater: | |
CII | Capaldi & Patterson (1989) | Not limited | Parent–child interaction | Parent behavior Child behavior | Parent warmth Appropriate discipline Harsh discipline Physical discipline Child warmth toward parent Child noncompliance toward parent | Coded for general impression after an interaction observation | Internal consistency: α = 0.84–0.91 for parenting behavior; α = 0.67 for child positive affect (Hurlburt et al., 2013) |
DPICS-II | Eyberg et al. (1994) | 3–7 years | Parent–child interaction Child deviant behavior | Amount of parental control Demand on child’s compliance | 27 subscales on parenting behavior 25 subscales on child behavior | In each case 5-min warm up, 15-min child-led play, parent-led play, and clean-up session (without warm up) | Interrater: ICC = 0.70–0.80 for parenting behavior; ICC = 0.79–0.87 for child behavior (Hurlburt et al., 2013) |
MBQS | Pederson & Moran (1999) | 6–42 months | Maternal sensitivity | 90 items that assess features of maternal sensitivity | 2–4 hr free home observation | Retest (4 months): rtt = 0.71 for sensitivity Interrater: ICC = 0.71–0.72 (Cicchetti et al., 2006) | |
P/CIS | Farran et al. (1986) | 3–60 months | Parental involvement | Amount Quality Appropriateness | Physical involvement Verbal involvement Responsiveness Play interaction Teaching Control of child’s behavior Positive statements Negative statements Directives Goal setting Facilitating relationships among activities | 30-min free play in a natural setting | Interrater: r = 0.88–0.94 Internal consistency: α = 0.85–0.96 (Fagan & Dore, 1993) |
Parent–Child Interaction Procedure | Heller et al. (1998) | 12–60 months | Parent–child relationship | Parental behavior Child behavior | Emotional responsiveness Behavioral responsiveness Positive affect Irritability Withdrawal/indifference Aggression towards the child Positive affect Withdrawal/Indifference Irritability/anger Noncompliance Aggression towards the parent Persistence Enthusiasm for the task | Crowell Procedure including seven different tasks (overall 30–45 min): free play, clean up, bubbles, and four teaching tasks | Interrater: r = 0.71–0.84 (Robinson et al., 2009) |
PMRS | Brassard et al. (1993) | 5–9 years | Psychological neglect and abuse | Maltreating behavior Positive parenting behavior | Spurning Terrorizing Exploiting/Corrupting Denying emotional responsiveness Quality of instruction Supportive presence Respect for child’s autonomy Strategies for maintaining task Involvement Emotional response to task Mental status Touching Interaction: Experience of mutual pleasure Body harmonics | 15-min teaching task | Interrater: rtt = 0.72–0.94 Retest (2 weeks): Exact agreement between 46% (denying emotional response) and 100% (terrorizing; Brassard et al., 1993) |
Qualitative Ratings of Parent–Child Interactions | Cox (1997) | Not limited | Parenting | Positive regard for the child Negative regard for the child Sensitivity Disengaged Intrusiveness Flatness of affect | 30-min interaction: free play, clean up, teaching task | Internal consistency: α = 0.78 Interrater: | |
SOC-III | Cerezo et al. (1986) | Not limited | Parental interaction with the child Child’s behavior | Valence Frequency Duration Sequence | Interactional: Approach Instruction Compliance Opposition Noninteractional: No response Interactional: Approach Instruction Compliance Opposition Noninteractional: Play Work Time out Rule violation Nothing Compliant | 60-min free interaction in a natural setting | Interrater: |
SASB | Benjamin, 1974 | Not limited | Transitive (parental behavior) Intransitive (child behavior) Introjection | Warmth Degree of interdependence Warmth Degree of interdependence Warmth Degree of interdependence | Emancipate versus control Affirm versus blame Active love versus attack Protect versus ignore Separate versus submit Disclose versus sulk Reactive love versus recoil Trust versus wall-off Self-emancipate versus self-control Self-affirm versus self-blame Active self-love versus self-attack Self-protect versus self-neglect | Different kinds of joint and teaching tasks | Interrater: κ = 0.64–0.84 (Skowron et al., 2011) |
UCLA Parent–Child Coding System | Valeri et al. (2001) | Not reported | Parent–child interaction | Parental behavior Child behavior | Hostility Negative detachment Emotional control Support Positivity Hostility Positivity Oppositionality Demandingness Compliance | 13-min interaction task: teaching, planning, conflict | Interrater: ICC = 0.62 for parent behavior; ICC = 0.62 for child behavior (Lau et al., 2006) |