Measuring the Extent of Woman Abuse in Intimate Heterosexual Relationships: A Critique of the Conflict Tactics Scales

Walter DeKeseredy and Martin Schwartz


Large-scale survey data have played an important role in sensitizing the media, government officials, and members of the general public to the shocking extent of woman abuse in intimate, heterosexual relationships. Today, many North Americans view male-to-female violence in marriage, dating, and cohabitation as a major social problem (Kline, Campbell, Soler & Ghez, 1997). In fact, there have been calls for the end to statistical surveys. Some have argued that the battle for recognition has been won, while others question whether quantitative techniques can ever adequately capture the complex experience of being battered.

Much of the support for continued quantitative research comes from those who contend that accurate statistics are essential to motivate government agencies to devote more resources to the development of prevention and control strategies. Of course, statistics are never sufficient to accomplish this end. However, as feminist scholars Bart, Miller, Moran, and Stanko (1989) point out: ""The principal questions that organize policy efforts are ultimately quantitative - how many are there, who are they, where are they, how bad are the consequences, how much will it cost?"" (p. 433). Those who fund programs tend to respond better to empirical data. Dealing with these data, however, has led politicians, journalists and scholars to a series of questions on who is at fault in battering, and how much battering actually takes place in society.

Although the academic and feminist literature is filled with debates on whether these instruments are flawed, the best-known and most often used quantitative technique designed to obtain estimates of the extent of physical woman abuse has been the Conflict Tactics Scale (CTS), and more recently the CTS2. Certainly qualitative methodologists and feminist researchers have employed a wide variety of other techniques and measures, but no other measure has approached the widespread use of the CTS. The main objectives of this brief document are to: (1) critique these two measures; and (2) provide suggestions for enhancing the quality of survey data on woman abuse in intimate heterosexual relationships, the focus of the CTS measurement.

The CTS and CTS2

The CTS was developed originally in the 1970s by University of New Hampshire sociologist Murray Straus to study violence within families. By now the original or a modified CTS appears at the core of research reported in over 100 scientific journal articles and at least 10 North American books. Although the CTS may in various studies be given only to men or only to women, the most widely cited work involves administering the survey to both men and women in intact heterosexual family units (married or cohabitants). The instrument solicits information from both men and women about the ""conflict tactics"" used by both men and women. The CTS consists of eighteen items that measure three different ways of handling interpersonal conflict in intimate relationships: reasoning, verbal aggression (referred to by some researchers as psychological abuse), and physical violence. These items are ranked on a continuum from least to most severe, with the first ten describing tactics that are not physically violent and the last eight describing violent acts. The last five items, from ""kicked, etc."" to ""used a knife or a gun,"" make up the ""severe violence index.""

The type of ""conflict tactic"" used to measure violence that occurred in the past year (incidence) is generally introduced to the respondent with the following preamble. Note the ideological and factual assumptions embedded in this introduction, such as the notion that battery is the result of an ""argument.""

No matter how well a couple gets along, there are times when they disagree, get annoyed with the other person, or just have spats or fights because theyíre in a bad mood or tired or for some other reason. They also use many different ways of trying to settle their differences. Iím going to read some things that you and your (spouse/partner) might do when you have an argument. I would like you to tell me how many times Ö in the past 12 months [Read item] (Straus, 1990, p. 33).

Research has suggested that the CTS seems to be a reliable method of eliciting highly sensitive data on the least known sides of intimate heterosexual relationships. For example, in both Canada and the U.S., city-wide, provincial/state, and national representative sample surveys that have used the CTS show that annually at least 11 percent of North American women in marital or cohabiting relationships are physically abused by their male partners. Many social scientists consider CTS data ""probably the best available when it comes to estimating the incidence and prevalence of woman abuse in the population at large"" (Smith, 1987, p. 177). Yet, quite a large number of researchers have criticized the CTS for the following reasons:


Although many of these critiques have been widely voiced for more than a decade, few researchers who use the CTS seem aware of them. However, Straus, Hamby, Boney-McCoy, and Sugarman (1995) recently developed the CTS2 to address some of these criticisms. To meet the concern that the CTS may not elicit responses on a variety of injurious behaviors, it includes more physical and psychological abuse items (e.g., ""I called my partner fat or ugly""). To deal with the strong attack that the CTS does not measure sexual violence, the CTS2 measures seven types of sexual assault. Finally, to allow researchers to tell the difference between events that cause physical injury and those that do not (e.g., slaps that break teeth, and those that might not cause physical injury), the CTS2 includes several injury or physical outcome measures, such as ""I needed to see a doctor because of a fight with my partner."" All of these are positive revisions that speak directly to some of the earlier criticisms.

Still, the CTS2 does not resolve all of the problems with the CTS. The most important place where the CTS2 does not improve on the CTS is that it continues to only situate abuse in the context of settling disputes or conflicts (the preamble remains the same). As suggested above, this limitation in effect tells the respondent to exclude reporting on abuse that is control-instigated or which does not arise from a known cause. It also does not allow the researcher to separate out aggressive abuse, whether physical or psychological, from those assaults used in self-defense.

Context, Meanings, and Motives Measures

Why do men and women use physical violence in marriage, dating, and other intimate, heterosexual relationships? As suggested above, the CTS does not provide adequate answers to this question. Much worse is that many people think that the answers the CTS provides do in fact deal with this question. The data that arise from the use of the CTS are commonly, and problematically, used to show that violence in relationships is ""sexually symmetrical"" (Dobash, Dobash, Wilson, & Daly, 1992). In other words, by simply counting the number of blows struck, the data appear to show that women are just as, if not more, violent than men. Unfortunately, this crude methodology can hide as much or more than it can illuminate (Schwartz & DeKeseredy, 1993).

These problems can be avoided by including questions about motives, meanings, and contexts in different sections of the CTS or CTS2. For example, DeKeseredy and Kelly (1993) placed the following three questions after both the first three and the last six violence items in the CTS, as part of a national study to measure the prevalence of violence in Canadian university and college dating:

On (the following) items, what percentage of these times overall do you estimate that in doing these actions you were primarily motivated by acting in self-defense, that is protecting yourself from immediate physical harm?

On (the following) items, what percentage of these times overall do you estimate that in doing these actions you were trying to fight back in a situation where you were not the first to use these or similar tactics?

On (the following) items, what percentage of these times overall do you estimate that you used these actions on your dating partners before they actually attacked you or threatened to attack you?

In analyzing the data generated by these questions, DeKeseredy et al. (1997) did not find support for the sexual symmetry thesis. Rather, a substantial number of women reported that their violence was in self-defense or ""fighting back."" These findings are consistent with Saundersí (1986) study of battered women. Thus far, the sexual symmetry thesis has only been supported by those using crude measures, such as the CTS with no further questioning.

The most important point of this paper is that the bulk of the research in this field has simply counted blows (who hit whom, and how often). The CTS2 speaks to one context issue (but only one) by asking about injury. A light slap may be different than one that jars loose several teeth. A push out of the way is different than a push down a flight of stairs. However, the survey still does not easily differentiate between a victim fighting back for her life, a survivor retaliating, and an instigator of violence without cause. All are considered violent. Even the more recent strategy of asking who struck the first blow (purportedly to tell who is the aggressor and who is fighting in self-defense) can be hard to place in context. When a woman has been beaten 30 times in the past and knows from her husbandís behavior that a beating is coming within minutes, and further knows that if she strikes first she will end up being hurt less, does that mean that the violence is the womanís fault?

Thus, both versions of the CTS have serious limitations. However, this does not mean that researchers should not use them, only that their studies will be flawed if they use the CTS or CTS2 as the sole measure of abuse. What are required, then, are multiple measures of abuse.

The Need for Multiple Measures of Woman Abuse

Although the use of multiple measures is a technique long recommended to enhance the reliability and validity in the measurement of social variables, most surveys of male-to-female abuse ignore this recommendation in favor of the simple use of some version or modification of the CTS. Unfortunately, regardless of how many abuse items respondents can choose from in either the CTS or CTS2, this unidimensional method of generating data does not provide respondents with sufficient opportunities to disclose abusive experiences. One method of gaining information is simply to ask for it. DeKeseredy and Kelly (1993), in their national survey on Canadian post-secondary dating, first asked:

Thinking about your entire university and/college career, have you ever been upset by dating partners and/or boyfriends trying to get you to do what they had seen in pornographic picture, movies, or books?

Those who answered ""yes"" were then asked to respond to the following supplementary question:

If you were upset, can you tell us what happened? Please provide this information in the space below.

This can be done for most questions. Leggett and Schwartz (1996), for example, asked a variety of questions that invited the respondent to go into more detail to explain her answer to multiple choice questions, or else to explain why she did not fit into any of their categories.

A different sort of multiple measure deals with the problem that in going through such a survey, people may not report incidents for several reasons, such as embarrassment, fear of reprisal, shame, or a reluctance to recall traumatic memories. However, several studies have shown that if respondents are asked again later by an interviewer or asked to complete self-reported, supplementary open- and closed-ended questions, some silent or forgetful participants will reveal in this second round having been victimized or abusive (DeKeseredy and Schwartz, 1998; Junger, 1990; Kelly, 1988). For example, Smith (1987) found that a substantial number of respondents changed their answers when asked the questions again in different words by a telephone interviewer. Belated responses increased the overall violence prevalence rate by approximately 10%. However, for the narrower set of question about severe violence, the prevalence rate increased by close to 60%, from 7.1% to 11.3%. It should also be noted that in addition to giving respondents more opportunities to disclose events, supplementary open-ended questions (where the respondent is given the opportunity to use her own words) build researcher-respondent rapport, allow respondents to qualify their responses, and overcome or minimize the hierarchical nature of traditional or mainstream survey research (Smith, 1994).


The CTS and CTS2, like other measures of woman abuse, have several strengths and limitations, and researchers have devoted substantial energy to either attacking or defending the empirical value of these techniques. Based on our own research experience and our review of the path-breaking research done by Smith, we contend that the CTS and CTS2 can contribute to the development of a rich data base on non-lethal forms of male-to-female assault. However, survey research that relies solely upon the CTS or only on one alternative measure tells just a part of the story. Such an approach contributes to massive underreporting and ignores the contexts, meanings, and motives of abuse.

Woman abuse is a multidimensional, complex problem. It warrants the use of multiple measures, in addition to measures asking about the specific context, meanings, and motives of respondents. Unfortunately, regardless of the methods used to generate woman abuse data, researchers will always have to face the fact that some respondents simply exercise their right to withhold information on abusive experiences. If there is a chance that abusers would overhear them or find out about disclosure, silence might be their best course of action, although that doesnít (from the researcher point of view) make for ""good data."" Indeed, perfect surveys on violence against women are not possible, but good ones can and should be done. The use of the CTS or CTS2 and supplementary open- and closed-ended questions can be a useful part of this process.

Authors of this document:
Walter S. DeKeseredy, Ph.D.
Carleton University, Ottawa

Martin D. Schwartz, Ph.D.,
Ohio University, Athens

February 1998

Distribution Rights: This Applied Research paper and In Brief may be reprinted in its entirety or excerpted with proper acknowledgement to the author(s) and VAWnet (, but may not be altered or sold for profit.

Suggested Citation: DeKeseredy, W. & Schwartz, M. (1998, February). Measuring the Extent of Woman Abuse in Intimate Heterosexual Relationships: A Critique of the Conflict Tactics Scales. Harrisburg, PA: VAWnet, a project of the National Resource Center on Domestic Violence/Pennsylvania Coalition Against Domestic Violence. Retrieved month/day/year, from:

(Items marked with an asterisk are recommended Key Readings)

Bart, P.B., Miller, P.Y., Moran, E., & Stanko, E.A. (1989). Editorsí introduction. Gender & Society, 3, 431-436.

Chang, V.N. (1996). I just lost myself: Psychological abuse of women in marriage. New York: Praeger.

DeKeseredy, W.S. (1995). Enhancing the quality of survey data on woman abuse: Examples from a national Canadian study. Violence Against Women, 1, 158-173.

DeKeseredy, W.S., & Kelly, K. (1993). The incidence and prevalence of woman abuse in Canadian university and college dating relationships. Canadian Journal of Sociology, 18, 137-159.

DeKeseredy, W.S., Saunders, D.G., Schwartz, M.D., & Alvi, S. (1997). The meanings and motives for womenís use of violence in Canadian college dating relationships: Results from a national survey. Sociological Spectrum, 17, 199-222.

*DeKeseredy, W.S., & Schwartz, M.D. (1998). Woman abuse on campus: Results from the Canadian national survey. Thousand Oaks, CA: Sage.

*Dobash, R.P., Dobash, R.E., Wilson, M., & Daly, M. (1992). The myth of sexual symmetry in marital violence. Social Problems, 39, 71-91.

Ellis, D., & Stuckless, N. (1996). Mediating and negotiating marital conflicts. Thousand Oaks, CA: Sage.

Junger, M. (1990). The measurement of sexual harassment: Comparison of the results of three different instruments. International Review of Victimology, 1, 231-239.

Kelly, L. (1987). The continuum of sexual violence. In J. Hanmer & M. Maynard (Eds.), Women, violence and social control (pp. 46-60). Atlantic Highlands, NJ: Humanities Press International.

Kelly, L. (1988). Surviving sexual violence. Minneapolis: University of Minnesota Press.

Kline, E., Campbell, J., Soler, E. & Ghez, M. (1997). Ending domestic violence: Changing public perceptions/halting the epidemic. Thousand Oaks, CA: Sage.

*Kirkwood, C. (1993). Leaving Abusive Partners. Newbury Park, CA: Sage.

Leggett, M. & Schwartz, M. (1996). Emotional reactions and blame in the sexual assault of women on the college campus. Paper presented at the annual meeting of the Academy of Criminal Justice Sciences, Las Vegas, NV, (March).

Saunders, D.G. (1986). ""When battered women use violence: Husband-abuse or self-defense? Violence and Victims, 1, 47-60.

Schwartz, M.D., & DeKeseredy, W.S. (1993). The return of the ""battered husband syndrome"" through the typification of women as violent. Crime, Law and Social Change, 20, 249-265.

Schwartz, M.D., & DeKeseredy, W.S. (1997). Sexual assault on the college campus: The role of male peer support. Thousand Oaks, CA: Sage.

Smith, M.D. (1987). The incidence and prevalence of woman abuse in Toronto. Violence and Victims, 2, 173-187.

Smith, M.D. (1994). Enhancing the quality of survey data on violence against women: A feminist approach. Gender & Society, 18, 109-127.

*Straus, M.A. (1990). Measuring intrafamily conflict and violence: The Conflict (CT) Scales. In M.A. Straus and R.J. Gelles (Eds.), Physical violence in American families (pp. 29-47). New Brunswick, NJ: Transaction Publishers.

Straus, M.A., Hamby, S.L., Boney-McCoy, S., & Sugarman, D.B. (1995). The revised Conflict Tactics Scales (CTS2-form A). Durham, NH: Family Research Laboratory.

* The production and dissemination of this publication was supported by Cooperative Agreement Number U1V/CCU324010-02 from the Centers for Disease Control and Prevention. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the CDC, VAWnet, or the Pennsylvania Coalition Against Domestic Violence."

Distribution Rights

This Applied Research paper and In Brief may be reprinted in its entirety or excerpted with proper acknowledgement to the author(s) and VAWnet, a project of the National Resource Center on Domestic Violence, but may not be altered or sold for profit.