Multivariate analysis
The final analysis combined all variables examined above in a multivariate logistic regression analysis, with student status as the dependent variable. This analysis demonstrated that the two types of students differed significantly on two items: medical students were more likely to be occasional smokers and less likely to report that smoking does no harm. While the univariate analysis showed that medical students were less likely to agree with the statement ‘smoking a few cigarettes every now and then has only benefits and does no harm’ (24.8% versus 29.6%, respectively), this did not reach statistical significance in the multivariate analysis.