Subject: Multivariate model
Details: I need to know how to report the findings from this type of model
How to report the results of a study that examines the association between driving patterns and medical conditions. Several conditions reached significance but were no longer significant in multivariate analysis. Is it necessary to report all of the findings that didn't reach significance in the multivariate model?
What about the other studies that reported that these conditions were significant but didn't use a multivariate model?
Would it be correct to say "The odds of decreasing mileage increased for individuals who had experienced falls (OR = 1.30 95% CI = 1.01 - 1.68)" without mentioning the other independent variables in the model?
Here's what the article says:
Means and standard deviations were computed for continuous variables. Proportions were calculated for categorical variables, and the differences were tested by chi-square test. The impact of individual medical conditions on driving variables was examined by logistic regression. Odd ratios from logistic regression models were presented to estimate the strength of the association. Significant odds ratios were entered into the multivariate model. A stepwise technique was used to select the final multivariate models with age, education, and residence controlled in the models.