As was observed in the current smoking, gender was not significant but grade and race/ethnicity were significant in predicting students' established smoking status in IYTS 2000 and 2004 (Tables 5 and 6).
As was observed in current smoking, exposure to pro-tobacco messages was a significant predictor for established smoking, whereas exposure to anti-tobacco messages and information about health consequences were not significant predictors in IYTS 2000 and 2004 adjusted logistic models.
This particular analysis of the Indiana Youth Tobacco Survey (IYTS) data was part of a larger study regarding youth tobacco use.
For example, in IYTS 2000, the age categories "18 years old" and "19+ years old" were merged into "18+ years old" because the frequency for "19+ years old" was too small (n = 6).
The statistical package SUDAAN (Research Triangle Institute, Research Triangle Park, NC) was used to estimate sampling variances and standard errors, which takes into account the complex sample design of IYTS. Sampling weights were used to adjust for unequal probabilities of selection, nonresponse, and disproportionate selection of different population groups.
The findings of this study support the inclusion of items in IYTS 2004 that measured perceived peer acceptance to smoking as they were significantly associated with OFS.