Make sure you have a primary exposure/predictor variable listed!

Make sure you have a primary exposure/predictor variable listed!

by Mark Pletcher -
Number of replies: 0

Hi DCR - I want to correct a mistake I've now seen several times on Protocol Outlines from Assignment #2.

The Protocol Outlines I saw were observational studies, and they wrote: "Intervention - None" (which is correct), but they did not replace "Intervention" with some other exposure.  I'm sure this is because the example I provided was a randomized controlled trial that DID have an intervention.  And it's true that your study, if it's not a trial, will not have an intervention.  HOWEVER, your study SHOULD still have an exposure variable of some sort.  The only exception would be if you are conducting a simple descriptive study (e.g., a simple survey of your whole target population on a single outcome variable).

Please make sure that your Protocol Outline has a primary exposure/predictor variable listed!  (Or be prepared to argue why it's a simple descriptive study).  You'll need this figured out to complete the Sample Size calculation for Assignment #3.