Measurement of Children's Exposure to Violence: A Rasch Analysis

• NCJ Number: 165450
• Journal: International Journal of Methods in Psychiatric Research Volume: 6 Issue: 161 Dated: (1996) Pages: 1-8
• Author(s): D J Kindlon; B D Wright; S W Raudenbush; F Earls
• Date Published: 1996
• Length: 8 pages
• Grant Number(s): 93-IJ-CX-K005
• Agencies: NIJ-Sponsored
• Publication Type: Report (Study/Research)

Annotation

Exposure-to-violence (ETV) response data were obtained from a demographically diverse sample of 49 subjects (29 females and 20 males) who ranged in age from 9 to 24 years old, so as to examine their fit to the Rasch model item response model for rating scales.

Abstract

Subjects were administered the Exposure to Violence Interview (Taylor et al., 1994) in a standard protocol along with other measures as part of a pilot test for a larger study. The response data were subjected to Rasch analysis. Rasch model constructs a linear measure for exposure to violence by using ordinal response data as realizations of the probabilities of item endorsement given the level of item extremity and the exposure level of the person taking the item. The resulting person measures and item calibrations are on the same scale. Rasch analysis proceeds in several steps; the first is to see whether the unconditional maximum likelihood interactive estimation converges quickly to a point where residuals are minimal. For these data, convergence occurred quickly with small measure residuals. This implies that the data were easy to fit into the Rasch model parameterization. The next step in determining the goodness of fit of the data to the model is to examine the distribution of persons and items on the linear logit scale. The results show that a psychometrically sound linear scale for the measurement of exposure to violence can be constructed by a Rasch model analysis. The hierarchical ordering of item difficulty was consistent with expectations and thus provides some evidence for construct validity of the scale. The experience with exposure to violence suggests that item response models such as Rasch may be applied to other multi-item assessments of social behavior and experience. 4 tables and 26 references