Item Response Theory Analysis of Intimate-Partner Violence in a Community Sample

University Behavioral Associates, Montefiore Medical Center, 334 East 148th Street (2nd Floor), Bronx, NY 10451, USA.
Journal of Family Psychology (Impact Factor: 1.89). 02/2012; 26(2):198-205. DOI: 10.1037/a0027100
Source: PubMed


The Revised Conflict Tactics Scale (CTS2; Straus, Hamby, Boney-McCoy, & Sugarman, 1996) is a widely used measure of physical, psychological, and sexual aggression as well as injury and negotiation between partners. In this study, we analyzed male-to-female and female-to-male physical aggression using data from 453 community couples. We used item-response theory (IRT) to assess the range and precision of physical aggression severity in the past year, as evaluated by the CTS2. Our analyses support a single dimension of physical aggression with quantitative, rather than qualitative, differentiation between minor and severe physical aggression items. Surprisingly, test information curves revealed that male perpetration and victimization items provided as much or more information as their counterpart female victimization and perpetration items over the same range of aggression severity. Finally, the data suggests that CTS2 items best assessed moderate-to-severe levels of physical aggression in the previous 12 months. However, virtually no information was assessed by male or female reports of perpetration or victimization items below or around the mean of aggression (i.e., theta; θ = 0). Suggestions for improved item coverage and implications for the assessment of aggression are discussed.

Download full-text


Available from: K. Daniel O'Leary, Jul 14, 2014
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Objective: Gender patterns in intimate partner violence (IPV) remain a controversial topic. Some self-report measures produce gender “parity” in IPV rates. However, other self-report surveys do not produce gender parity, nor do arrests, reports to law enforcement, homicide data, helpseeking data, or witness reports. This methodological inconsistency is still poorly understood. The objective of these studies is to explore the effects of item wording on gender patterns for victimization reports in a range of samples. Method: In Study 1, 238 undergraduates were randomly assigned either the standard Conflict Tactics Scales (CTS) physical victimization items or a version which changed the partner-specific wording to generic wording (“Someone” instead of “My partner”), with perpetrator information collected in follow-up. Studies 2 and 3 compared the standard approach to items with stems intended to reduce false positives (either “Not including horseplay or joking around . . .” or “When my partner was angry . . .”), among 251 college students and 98 agency-involved women, respectively. Study 4 implemented the “not joking” alternative from Study 3 in a large rural community sample (n = 1,207). Results: In Studies 1 and 2, significant Wording × Gender analyses indicated that some item wordings yielded higher rates of female than male victimization. Study 3 showed similar patterns across forms for highly victimized women. Study 4 found higher female than male victimization for a new scale and every item. Conclusion: The CTS and similar behavioral checklists are unusual in their inattention to false positives. Self-report measures designed to minimize false positives produce results consistent with other IPV methodologies; that is, they do not demonstrate gender parity. The Partner Victimization Scale, described here, can be used when a scale that has multimethod convergence with other IPV methodologies is desired.
    Psychology of Violence 01/2014; DOI:10.1037/a0038207 · 1.83 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The analysis of patient-reported outcomes or other psychological traits can be realized using the Rasch measurement model. When the objective of a study is to compare groups of individuals, it is important, before the study, to define a sample size such that the group comparison test will attain a given power. The Raschpower procedure (RP) allows doing so with dichotomous items. The RP is extended to polytomous items. Several computational issues were identified, and adaptations have been proposed. The performance of this new version of RP is assessed using simulations. This adaptation of RP allows obtaining a good estimate of the expected power of a test to compare groups of patients in a large number of practical situations. A Stata module, as well as its implementation online, is proposed to perform the RP. Two versions of the RP for polytomous items are proposed (deterministic and stochastic versions). These two versions produce similar results in all of the tested cases. We recommend the use of the deterministic version, when the measure is obtained using small questionnaires or items with a few number of response categories, and the stochastic version elsewhere, so as to optimize computing time. Copyright © 2015 John Wiley & Sons, Ltd. Copyright © 2015 John Wiley & Sons, Ltd.
    Statistics in Medicine 03/2015; 34(16). DOI:10.1002/sim.6478 · 1.83 Impact Factor