Why do we estimate and report unadjusted regression results besides adjusted?

[Jungran Lee]

Hi All,

I hope you are well!
I am wondering why do we have to estimate and report unadjusted regression results.

I guess adjusted regression results are more precise, so I think those results are just enough to report.
But, in research articles, researchers usually include unadjusted estimates as well.
What reasons are there reporting unadjusted regressions??

Thank you for your answer in advance

[Clyde Schechter]

I'm not aware of any rule or custom that says you must always report unadjusted estimates as well as adjusted ones. It depends on the situation. Sometimes the unadjusted analyses are of interest in their own right. While we often think of adjustment as necessary to control for omitted variable bias, sometimes the model with the omitted variable is the one that answers the research question and the adjusted one does not. It all depends on what the research question is and how the results will be understood and applied.

[Rich Goldstein]

Clyde Schechter I would expand briefly on Clyde's answer: it will also depend on the design; in particular, a randomized control trial (if designed and implemented well) does not need any adjustment but may gain statistical power thru adjustment