2.3. Statistical Analysis

The two outcomes of interest for this study were whether allied health professionals provided: (1) dietary advice and (2) advice on increasing physical activity. Health professional and workplace characteristics and potential confounders (gender, highest level of qualification, area of practice, location of practice, years in clinical practice, hours worked per week, number of new clients per week, proportion of clients overweight or obese, number of referrals made to other services for weight management, department guidelines for providing weight management advice, having received education in weight management during their entry-level professional qualification, and continuing education in weight management) were compared across types of health professional groups using Pearson’s chi-square test and analysis of variance. Percentages were used to indicate the proportions of health professionals providing dietary and physical activity advice in various formats (i.e., printed or online material, or client discussion). Logistic regression models were used to determine if health professional group, attitudes, and workplace factors were associated with the provision of (1) dietary or (2) physical activity advice. All variables of interest were initially included in the models, and backwards stepwise elimination was used to exclude variables which were not significant at the p = 0.1 level based on the Wald Test. Odds ratios with 95% confidence intervals (CI) are reported. The statistical analysis package STATA V11.1 (StatCorp, College Station, TX, USA) was used for analysis [25].