Go to the Table Of Contents

Worker Drug Use and Workplace Policies and Programs: Results from the 1994 and 1997 National Household Survey on Drug Abuse

Previous PageTable Of ContentsList Of FiguresList Of TablesNext Page

CHAPTER 6: MULTIVARIATE MODELS OF CURRENT ILLICIT DRUG USE AND CURRENT HEAVY ALCOHOL USE AMONG FULL-TIME WORKERS



3 Directly interpreting the effect of each explanatory variable in a logistic regression requires interpreting the outcome variable as a logarithm of the odds of two probabilities, rather than one probability. The natural logarithm of the odds, log {P(Y=1)/[1-P(Y=1)]}, is the logit of Y. The exponential regression coefficients are reported here. These values represent odds ratios (ORs) and may be interpreted as the relative likelihood {P(Y=1)/[1-P(Y=1)]} of current illicit drug use or current heavy alcohol use with each unit increase in the predictor variable, controlling for the other variables in the model.

4 Magnitudes of positive and negative effects can be compared by taking the inverse of the negative effect (or vice versa). For example, the magnitude of effect indicated by a positive OR of 2 is the same as the magnitude of the effect indicated by a negative OR of .5 (=2).

Previous PageTop Of PageTable Of ContentsList Of FiguresList Of TablesNext Page

This page was last updated on December 30, 2008.