Measuring gender differences in partner violence: implications from research on other forms of violent and socially undesirable behavior

Sex Roles: A Journal of Research,  June, 2005  by Sherry L. Hamby

• E-mail
• Print
• Link

In particular, criminologists have demonstrated that self-report measures identify less serious behavior, on average, than official records do (Hindelang et al., 1979; Menard & Covey, 1988). Many acts disclosed in self-report measures would not normally be classified as crime. Even items that appear to refer to serious infractions often lead to reports of trivial offenses that do not meet the legal definitions of the crimes they were intended to measure, amounting to 20-36% of all reports across several studies (Cohen & Land, 1984; Hindelang et al., 1979; Huizinga & Elliott, 1986; Levine, 1976). For example, a youth described as above the 90th percentile in delinquency in one sample reported no physical violence, but incidents such as helping himself to beers from a friend's refrigerator and lying about his age to get a reduced ticket price at the movie theater (Gold, 1970, cited in Hindelang et al., 1979). Although these behaviors are not laudable, Hindelang and his colleagues questioned whether they constituted what most people mean by serious delinquency. Similarly, Garofalo, Siegel, and Laub (1987) found that some minor incidents were counted in NCVS/NCS crime rates. For example, they questioned whether it was appropriate to classify a child being hit on the head with a lollipop as an aggravated assault.

Many self-report measures of crime and delinquency include numerous minor offenses among their items (Huizinga & Elliott, 1984). This is due in part to statistical needs to identify a minimum percentage of individuals as "violent" or "delinquent." Similarly, most partner violence inventories include grabbing and other physical aggression that would seldom be deemed a crime by police (no matter the victim-offender relationship), and sulking or other psychological aggression that may not meet the typical social construction of "abuse." Minor forms of aggression are typically the most frequently endorsed in partner violence surveys. Self-report checklists are probably capturing different phenomena than crime statistics.

Many Incidents are Ambiguous

Not only do self-report surveys tend to collect more minor incidents, they may also collect some incidents that would not normally be construed as violence at all, or are ambiguously aggressive. In a comparison of CTS reports with other self-report measures of partner violence victimization, one-third of a sample of college women were classified inconsistently across three measures--that is, they were classified as victims on at least one measure but not on all three (Hamby et al., 1996). Some misclassifications were due to serious incidents, such as being thrown across the room, that are not represented on the CTS. Most, however, were because someone who reported being pushed, shoved, or grabbed, did not identify that behavior as violence on another measure. If only respondents who reported violence across all three measures were counted as victims, then the rate of victimization in that study was halved.