The estimators, which use data from the National Crime Survey, were developed by Eddy, Fienberg, and Griffin; Griffin; and Alexander. The ad hoc estimator of Eddy, Fienberg, and Griffin was consistent only under a model which assumed that the probability of reporting a victimization did not depend on the number of months of information that a housing unit contributed during the year being considered. A modified version of this estimator was consistent under a model which appeared, at least for a small sample, to fit the data well. The models under which the estimators of the Bureau of Justice Statistics and Griffin were consistent were similar to the model for the ad hoc estimator in that the probability of victimization in half of the year had to be the same as the probability of victimization in a full year. Figures and 16 references are provided. (Author summary modified)
Downloads
Similar Publications
- Cognitive and contextual influences in determination of latent fingerprint suitability for identification judgments
- Insights into turning points from the perspective of young people with out-of-home care experience: events, impact and facilitators of change
- Improving Prison Classification Procedures in Vermont: Applying an Interaction Model, 1975-85