NCJ Number
65927
Date Published
1979
Length
324 pages
Annotation
TO EXPLORE THE ROLE OF RESPONSE ERRORS IN SELF-REPORTED MEASURES OF CRIME, THIS DISSERTATION CENTERS ON THE LEVELS AND EFFECTS OF RESPONSE ERRORS IN THE CONTEXT OF AN EMPIRICAL STUDY OF REPORTED OFFENSES.
Abstract
AS BACKGROUND TO THE STUDY, THEORIES OF MEASUREMENT AND STRUCTURAL MODELS OF SURVEY RESPONSES ARE DISCUSSED, AND MODELS ARE CONTRASTED FOR THEIR TREATMENT OF ERRORS IN ONE-ITEM AND MULTIPLE-ITEM TESTS. THEN A STRUCTURAL MODEL OF REPORTED OFFENSES IS PROPOSED FOR EXAMINING TREATMENTS OF ERROR IN PREVIOUS SELF-REPORT STUDIES OF CRIME. USING THE STRUCTURAL MODEL OF RESPONSES, THE TREATMENT CENTERS ON THE TYPES AND PATTERNS OF ERROR OBSERVED IN THIS RESEARCH. THE GRAVITY OF THE RESPONSE-ERROR PROBLEM, THE VARIATION IN TYPES OF ERROR, AND THE OBSERVED CORRELATES OF ERROR ARE EXAMINED. SUBSEQUENTLY, DISCUSSION SHIFTS TO AN EMPIRICAL STUDY IN WHICH RESPONSE ERRORS ARE OPERATIONALIZED IN TERMS OF OFFENSES RESULTING IN ARRESTS. ESTIMATES OF RESPONSE ERROR ARE OBTAINED FROM COMPARISONS OF SURVEY RESPONDENTS' REPORTED AND OFFICIAL ARREST HISTORIES. BECAUSE RESPONSE ERRORS MAY DISTORT AGGREGATE AS WELL AS INDIVIDUAL ESTIMATES OF CRIME, THE AGGREGATE LEVELS OF RESPONSE BIAS IN SELF-REPORTS OF ARREST ARE EXPLORED; EMPIRICAL FINDINGS ARE PRESENTED. THE DIRECT RELATIONSHIP BETWEEN ESTIMATES OF RESPONSE ERROR AND REPORTS OF HIDDEN CRIME IS ALSO CONSIDERED, ALONG WITH HOW THIS RELATIONSHIP VARIES IN REPORTING DIFFERENT TYPES OF OFFENSES. FINALLY, STUDY FINDINGS ARE SUMMARIZED, IMPLICATIONS OF SUCH FINDINGS FOR ONGOING AND FUTURE RESEARCH ARE REVIEWED, AND RECOMMENDATIONS ARE MADE ON HOW THE EFFECTS OF SURVEY RESPONSE ERRORS MIGHT BE COUNTERBALANCED TO MORE ACCURATELY ESTIMATE TRUE PATTERNS AND TRENDS IN CRIMINALITY. AN APPENDIX CONTAINS THE FOLLOWUP INTERVIEW SCHEDULE. TABULAR DATA, CHARTS, FOOTNOTES, AN INDEX, AND A BIBLIOGRAPHY ARE INCLUDED. (WJR)