MEASUREMENT INSTRUMENTS IN CHILD ABUSE RESEARCH

• NCJ Number: 142815
• Author(s): M A Straus
• Date Published: 1993
• Length: 12 pages
• Sponsoring Agency: US Dept of Health and Human Services
• Sale Source: University of New Hampshire, Family Research Laboratory
• Publication Type: Research (Applied/Empirical)
• Language: English
• Country: United States of America
• Note: Revised version of paper prepared for the National Academy of Sciences, National Research Council, panel on Research on Child Abuse and Neglect, 1992

Annotation

Data were obtained from research studies of child abuse and neglect to assess the extent to which instruments meeting minimal psychometric standards were used in child abuse research and to examine the variables measured.

Abstract

A total of 714 articles were reviewed, and a measurement instrument was defined as a procedure combining values of three or more items or indicators to gauge an underlying continuum which could only be partly measured by a single item. Instruments were classified into one of four categories: maltreatment, effects of maltreatment, risk factors or causes of maltreatment, and other. Of the 714 articles, 137 articles used 83 different instruments, with 24 instruments used in two or more studies; 81 percent of the articles used no instrument to measure a variable. The least frequently measured phenomenon was child abuse itself; 41 percent of the instruments were intended to identify risk factors or causes of maltreatment, and 38 percent measured hypothesized effects of abuse. The findings suggest that research on child abuse has made minimal use of multi-indicator measures with known reliability and validity. Reasons for the limited use of standard measures are noted, and steps to improve measurement are discussed. Data on the instruments used between 1979 and 1989 are appended. 11 references and 1 figure