StATS: Presenting unadjusted and adjusted estimates side by side (created 2008-03-24).

This page is moving to a new website.

Someone on the Medstats discussion group asked about reporting the analysis of a model without adjustment for covariates along with the analysis adjusted for covariates. What is the purpose of reporting the unadjusted analysis?

I like to see both analyses because it lets you know whether the adjustment for covariates has had any practical impact.

Also, there is a pragmatic consideration. The unadjusted analysis represents a value that typically can be calculated by hand. In a logistic regression model comparing two groups, for example, the unadjusted odds ratio can be calculated directly from the 2 by 2 table. I like to be able to double check a few of the numbers presented in a paper just to get comfortable with the results.

Finally, the simplicity of the unadjusted estimate (if it is not seriously biased) further reinforces the credibility of the research. There's a general perception among some cynics that if you used a complicated statistical model, it was only because the simple model did not produce results you liked. When you show that the simple model produces the same results, it takes that argument away from your critics.

This page was written by Steve Simon while working at Children's Mercy Hospital. Although I do not hold the copyright for this material, I am reproducing it here as a service, as it is no longer available on the Children's Mercy Hospital website. Need more information? I have a page with general help resources. You can also browse for pages similar to this one at Category: Modeling issues, Category: Writing research papers.