The essence of synthetic estimation as applied to drug issues is the use of a presumed or established functional relationship, based partly on what is observable, to infer drug use when direct observations are unavailable. The author briefly reviews recent studies in which synthetic estimation, together with other estimation procedures, was used to estimate the prevalence of cocaine and heroin use. He then discusses a specific real-world application of the synthetic estimation approach that was developed for purposes of policy research. This application yielded estimates of approximately 2 million weekly users of cocaine and 660,000 weekly users of heroin in the United States during 1990. The author advises that in practice the application of synthetic estimation to drug use prevalence is more art than science. The data from the calibration sample are typically sparse, biased, and otherwise inaccurate. All three conditions mitigate against making probability-based estimates. Still, when used in concert with other complementary estimation techniques, it is useful in providing an information base for policymaking. 1 table, 19 notes, and 45 references
Downloads
Related Datasets
Similar Publications
- Dirichlet-Gamma-Poisson Model of Repeated Events - A Multivariate Description of Criminal Victimization in American Cities
- Drug Use as a Predictor of Rearrest or Failure to Appear: A User's Guide to the Machine-Readable Files and Documentation, Original Instruments, and Codebook
- Report From the States on What Works at the State and Local Levels: A Compendium of Assessment and Evaluation Results