Markov Chains for Random Urinalysis II: Age-Test Model with Absorbing State
Abstract
This is the second of a series of papers on the use of Markov chains to model random urinalysis programs. Previous work introduced the age-test Markov chain. This chain was used to model random urinalysis strategies stratified by time since last tested. This paper extends the age-test model by including an absorbing state for detection of drug users. We quantify the extent to which age-test urinalysis strategies can be gamed by drug users. The theory and formula for calculating the distribution of time to absorption (e.g., detection of drug users), the mean time to absorption and the expected number of visits to intermediate states (e.g., number of tests) prior to detection are developed. This model is used to analyze three age-test strategies.... Drug urinalysis, Markov chains, Stratified random sampling, Age-test model.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1993
- Accession Number
- ADA265557
Entities
People
- Douglas J. Hentschel
- James P. Boyle
- Theodore J. Thompson
Organizations
- Bureau of Naval Personnel