Paul E. Tracy

Randomized response: a method for sensitive surveys.

Survey research is useful in studying a variety of problems but asking respondents direct questions about controversial topics may not lead to honest answers. Is it possible to get good estimates of behavioral frequencies when a behavior is severely disapproved of and perhaps even criminal? One can get honest answers but not by asking naive questions. This monograph describes an innovative survey method designed for sensitive or threatening inquiries. The 1st part of the monograph discusses error sources especially in responses to sensitive disapproved or criminal behaviors and attitudes. The authors consider traditional strategies for protecting anonymity and confidentiality and note that these protections are either limited or perceived by respondents to be ineffective. The monograph focuses on describing the randomized response method and how it can be used to effectively and efficiently protect respondents thus minimizing survey bias. Randomized response can be used to 1) estimate parameters of both qualitative and quantitative measures 2) test subgroup differences and 3) perform bivariate and multivariate analysis. The authors consider survey design and include sample questions and research examples. An appendix contains sample instructions to respondents.

Continuity and Discontinuity in Criminal Careers

Introduction. Theoretical and Policy Considerations. Reviews of Related Research. Research Design and Methodology. The Prevalence of Adult Criminality. Predicting Adult Crime Status from Deliquency, Age, Frequency, and Court Dispositions. Predicting Adult Crime Status from Type of Deliquency. Predicting Adult Crime Status from the Severity of the Deliquency Career. Summary of Results. Implications for Theory and Policy. Index.

The Validity of Randomized Response for Sensitive Measurements

Randomized response is a survey technique for reducing response bias arising from respondent concern over revealing sensitive information. There has been some question whether bias reduction earned through the randomized response approach is sufficient to compensate for its inefficiency. By comparing self-reported arrests for two interview conditions (randomized response and direct question) with corresponding true scores appearing in police arrest files, a field-validation of a quantitative randomized response model was attempted. Overall, randomized response outperformed the more traditional direct-question method. Not only was there substantial reduction in mean response error, but the response error operative in the randomized response condition appeared to be random rather than systematic. A mean squared error comparison of the two conditions appears to assuage the concern over its relative inefficiency.

Gender Differences in Delinquency and Juvenile Justice Processing Evidence From National Data

This article traces the historical coverage of the gender issue in the criminological literature. It also provides contemporary empirical evidence about differences and similarities between girls and boys with respect to juvenile crime and to processing by the juvenile justice system, by analyzing several national juvenile crime data series, all of which have been recently updated. This research has produced numerous results that indicate that female and male delinquents possess more similarities than differences concerning arrest rates, arrest statistics, juvenile court data, and juvenile corrections data.

Serious, violent, and chronic juvenile offenders: The relationship of delinquency career types to adult criminality

We describe the intersections between serious, violent, and chronic juvenile offenders through 16 years in which the 1958 Philadelphia birth cohort was at risk for delinquency and young adult crime. Official records mark the offense careers from early onset as juveniles through age 26 for 27,160 persons. This large database is unique in its ability to support analysis of serious, violent, and chronic offending while controlling for differences based on race, gender, and neighborhood social status. The results show that belonging to certain delinquency subgroups or following certain “pathways” increases the likelihood of being arrested in adulthood. Although prevalence differed, general findings were consistent across demographic criteria. We demonstrate the importance, to developmental criminology, of linking juvenile delinquency career types to adult criminality for policy formulation and theory development.

The Randomized Response Approach

A considerable amount of attention has been afforded the randomized response approach over the past decade in an attempt to refine this method of reducing evasive answer bias in surveys of sensitive information. In this article we review these developments in an introductory fashion and address various concerns in designing a randomized response instrument. Last, the potential applicability of this approach to criminal justice research and evaluation is discussed.

Delinquency in Puerto Rico : the 1970 birth cohort study

Foreword, by Miguel Hernandez Agosto Background for the Current Study Prevalence Incidence Age and Delinquency Delinquent Recidivism Police and Court Dispositions Cohort Comparisons Summary Policy Implications and Suggestions for Legislation Appendices Bibliography Index

A measure of skewness in offense distributions

It has become a criminological fact of life that a small group of offenders is responsible for a far greater share of offending than its size would suggest. For example, it is well known that in the 1945 Philadelphia birth cohort, 6% of the cohort members committed 52% of the offenses. We argue, however, that the conventional percentaging approach to measuring skewness in offense distributions is inadequate. We propose and apply an alternative approach to measuring offense skewness which utilizes the offense distribution to its fullest and can easily and meaningfully be used for comparisons across or within study populations.

Adjusting for Unit Non-Response in Surveys Through Weighting

Unit non-response is a serious problem in survey research. This article validates the necessity of adjusting for unit non-response in disproportionate stratified sampling designs through the use of sample weights. Using data from the 1958 Birth Cohort study, we demonstrate that sample data which are affected by unit non-response can be a poor representation of population parameters. These non-response effects can be addressed through the application of sample weights.

Adjusting for Design Effects in Disproportionate Stratified Sampling Designs Through Weighting

This article validates the necessity of adjusting for the design effects in disproportionate stratified sampling designs through the use of sample weights. Using data from the 1958 Birth Cohort study, we demonstrate that complex sampling designs introduce sampling error and even sampling bias into sample data. Such sample data are a poor representation of population parameters. These design effects can be addressed through the application of sample weights.