CommentaryIdentification of Child Maltreatment With the Parent-Child Conflict Tactics Scales: Development and Psychometric Data for a National Sample of American Parents☆
Introduction
Most research on physical and psychological maltreatment of children, and virtually all research on neglect, is based on cases obtained from treatment or judicial agencies. Despite the importance of these clinically based studies, only a small fraction of maltreatment cases are known to social or judicial agencies. Moreover, there is evidence from studies of alcoholism and other social and psychological problems that cases in the general population suffering from the same problem may differ from “clinical” cases in ways that affect treatment or prevention programs (Straus 1990b). Consequently, epidemiological research on child maltreatment in the general population is needed in addition to clinically based studies.
One of the requirements for epidemiological survey research on the general population is a practical method of ascertaining the presence and degree of maltreatment. The Conflict Tactics Scales or CTS (Straus 1979, Straus 1990a; Straus et al. 1996) was designed to meet that need. This paper describes a new version of the CTS called the Parent-Child Conflict Tactics Scales (CTSPC). The CTSPC is intended to measure psychological and physical maltreatment and neglect of children by parents, as well as nonviolent modes of discipline.
The original CTS (to be called CTS1 from here on) has important limitations as a measure of child maltreatment (discussed in Straus and Hamby 1997). The limitations stem from the fact that the CTS1 was designed for use with partners in a marital, cohabiting, or dating relationship. To adapt it for measuring parental behavior, the main modification was to change the referent person from “your partner” to a specific child. Although the CTS1 has worked remarkably well as a measure of child maltreatment (see the review in Straus and Hamby 1997), some items were not really appropriate for parent-child relationships, and some important parental behaviors were not included.
A computer search for 1980 through 1996 revealed 132 publications that reported results from using the CTS1 to measure child maltreatment (see bibliography in Straus 1995). Most were based on data from responses by parents. A substantial number used the CTS1 to obtain recall data from adults about the behavior of their parents. Twenty-two studies were based on administration of the CTS1 to children ranging in age from 6 through 17. More important than frequency of use is the evidence accumulating from these studies of the concurrent and construct validity of the CTS1 as a measure of maltreatment of children (Straus and Hamby 1997; Yodanis et al. 1997).
An elementary but critically important indicator of validity is the fact that the rates of severe physical assault found by nine different investigators (summarized in Straus and Gelles 1990b, , part B) show that the CTS identifies many more cases than are known to child protective services (CPS). This is consistent with the long standing belief of CPS workers that there are many times more cases than are referred to them. A type of concurrent validity is the level of agreement between different members of the same family, such as the extent to which the report of a parent agrees with that of a child. Straus and Hamby 1997summarize the results of six such studies, each of which found substantial agreement. The most extensive evidence is on construct validity. Most of the studies located in our search provide evidence of construct validity because they report findings that are consistent with previously established empirical findings such as etiological links between physical abuse of children and stress (Eblen 1987) and depression (Campbell et al. 1997; Zuravin 1989) or between having experienced abuse and many kinds of maladaptive behavior such as delinquency and substance abuse (Miller et al. 1989), psychopathology (Dutton et al. 1996) and scores on the Child Abuse Potential Inventory (Caliso and Milner 1992).
The extent to which the CTS1 has been used in research on child maltreatment, despite its limitations, suggests that the revision described in this paper could make the CTSPC an even more useful instrument that the CTS1.
Section snippets
Theoretical Basis.
Despite important differences between the CTS1 and the CTSPC, the theoretical basis and mode of operationalization are fundamentally the same. The theoretical basis of the CTS is conflict theory (Adams 1965; Coser 1956; Dahrendorf 1959; Scanzoni 1972; Simmel 1955; Sprey 1979; Straus 1979). This theory assumes that conflict is an inevitable part of all human association, whereas physical assault as a tactic to deal with conflict is not (Coser 1956; Dahrendorf 1959). Consistent with this
The Gallup Survey
The data for this paper was obtained from a survey initiated and sponsored by the Gallup Organization as part of its National Social Audit Program (Gallup survey, #765). The authors of this paper all participated in the design of the questionnaire, including the design of the CTSPC. The interviews were pretested on a sample of 14 cases, and the wording of some items was revised to correct problems that were uncovered.
Prevalence and Chronicity
We will comment only briefly on the prevalence and chronicity statistics from the Gallup Survey because this is a methodological paper rather than a substantive report. But at least some consideration of these statistics is needed to evaluate the extent to which the CTSPC has been able to identify a meaningful pattern of parent behavior.
Reliability
A problem with the parent-child version of CTS1 that we hoped to remedy with the CTSPC is low internal consistency reliability. The alpha coefficients from seven analyses of the CTS1 data averaged .58 for the overall Physical Assault scale and .68 for the Psychological Aggression scale (Straus and Hamby 1997). The CTSPC alpha coefficients for the present sample of parents are similar: Overall Physical Assault scale = .55, Psychological Aggression = .60, Nonviolent Discipline = .70, and the new
Discriminant and Construct Validity
Evidence of construct validity occurs when a test is correlated with other variables for which there are theoretical or empirical grounds to expect an association (Campbell and Fiske 1959) and are not correlated when there are grounds for expecting the two variables to be uncorrelated. An overall judgment concerning construct validity takes many such linkages, including findings from different studies. An accumulated body of such evidence exists for the parent-to-child version of the CTS1 (see
Discussion
The Parent-Child Conflict Tactics Scales (CTSPC) is a revision of the Conflict Tactics Scales (CTS) that is specifically focused on parental behavior. The modifications include:
- •
Revision of the Psychological Aggression and Physical Assault scales to (1) improve clarity and age-appropriateness; (2) add items to increase content validity; (3) better differentiate between levels of severity of aggression by parents.
- •
Replacement of the Reasoning scale by a Nonviolent Discipline scale.
- •
New
Educational and Ethnic Group.
Although the CTSPC was designed to facilitate self-administration, it can be administered as a face-to-face or telephone interview. In revising the CTS we kept the vocabulary and sentence structure simple to make it applicable to a broad section of the population. The Flesch grade level measure (Flesch 1949) for the three basic scales (Nonviolent Discipline, Psychological Aggression, and Physical Assault) is 6th grade. If the supplemental scales (Neglect and Sexual Maltreatment) are included,
Acknowledgements
Acknowledgements—The authors wish to express their appreciation to the Gallup Organization for the survey to obtain the data reported in this paper.
References (63)
- et al.
Childhood history of abuse and child abuse screening
Child Abuse & Neglect
(1992) - et al.
Mother’s age and risk for physical abuse
Child Abuse & Neglect
(1992) - et al.
Sexual abused children in a national survey of parentsMethodological issues
Child Abuse & Neglect
(1997) Methodological issues and practical problems in conducting research on maltreated children
Child Abuse & Neglect
(1994)- et al.
Verbal aggression by parents and psychosocial problems of children
Child Abuse & Neglect
(1991) - Acock, A. (1997). Discussion of Straus et al. CTSPC article. Presented at the 1997 Theory Construction and Methodology...
Coercion and consensus theoriesSome unresolved issues
American Journal of Sociology
(1965)- et al.
Screening for physical violence in couple therapyMethodological, practical, and ethical considerations
Family Process
(1994) Psychological distress and the recall of childhood family characteristics
Journal of Marriage and the Family
(1991)AggressionIts causes, consequences, and control
(1993)
Convergent and discriminant validation by the multitrait-multimethod matrix
Psychological Bulletin
Predictors of depression in battered women
Violence Against Women
Men who batter womenA study in power
Journal of Family Violence
The functions of social conflict
The Beck Depression InventoryItem order and the impact of response sets
Journal of Personality Assessment
Class and class conflict in industrial society
System-initiated warrants for suspects of misdemeanor domestic assaultA pilot study
Justice Quarterly
Antecedents of abusive personality and abusive behavior in wife assaulters
Journal of Family Violence
Relative effectiveness of group treatments for men who batter
Social Work Research and Abstracts
The effects of group treatment for men who batterAn 18-month follow-up study
Research on Social Work Practice
The art of readable writing
Validity of using the Conflict Tactics scales in the Indianapolis prosecution experiment
Disciplining children in AmericaA Gallup Poll report
Child, maternal and family characteristics associated with spanking
Family Relations
Court responses to petitions for civil protection orders
Journal of Interpersonal Violence
Characteristics of perpetrators of family and nonfamily assaults
Hospital and Community Psychiatry
Correlates of early violence experience among men who are abusive toward female mates
Perceived social support and competence in abused childrenA longitudinal perspective
Journal of Family Violence
Children’s perspectives in the assessment of family violencePsychometric characteristics and comparison to parent reports
Child maltreatment
Cited by (0)
- ☆
The research was funded by National Institute of Mental Health grant T32MH15161 and the University of New Hampshire. This paper is a publication of the Measurement Research Program of the Family Research Laboratory, University of New Hampshire. A list of publications on measurement will be sent on request.