Measures of Association & Log Odds

Dear Statistics Folks,

The logistic regression is often misunderstood for several reasons. Some people think it’s just an extension of the linear regression and they interpret the parameter estimates like it’s all good. Others think all you have to do is note statistical significance of the beta estimates and which variables are significant to the model. What about your odds ratio? What about the probability for a particular statistic?

You have to include the odds ratios of your explanatory or independent variables in your interpretation of the statistical model. Because you want to know the odds of having Y=1, your dependent outcome variable, given your explanatory factors.

So once you obtain the parameter estimates, remember it represents the log-odds and not the true odds ratio. You must get the exponential of your model for each covariate in order to determine the odds in terms of each independent factor.

Your parameter estimates simply represent the log change in your Y ,dependent variable, given your X factor.So when people ask you what is your measure of association for the regression, you should interpret the  odds ratio.

If you enjoyed this blog there are more to follow.

 

Amy

 

 

 

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.