Skip to main content
. 2007 Jul;97(7):1311–1318. doi: 10.2105/AJPH.2005.071266

TABLE 3—

Parameter Estimates (B) for Covariate Effects in Discrete-Time Hazard Models of Age of First Marijuana Use

Regression Estimates
Ba (SE) Odds Ratio
Sample
    NP tribes –.26* (–.11) .78
    SW tribe –.90* (–.12) .41
Birth cohort × sample
    Younger, NP tribes .56* (–.11) 1.75
    Younger, SW tribe .74* (–.12) 2.09
Male gender .54* (–.07) 1.72
    Birth cohort × gender
    Younger males –.37* (–.08) .69
Sample × gender
    Males, NP tribes .01 (–.10) 1.01
    Males, SW tribe .39* (–.12) 1.47

Note. NP = Nothern Plains; SW = Southwest. n = 9492; number of events = 4468.

aMain effects Bs are the parameter estimates for the referent group, and interaction effects represent differential effects across birth cohorts, genders, and tribes. The SE corresponding to each parameter estimate (B) is the standard error of the estimate.

* P < .001