Timeline for How control for a pre-treatment outcome $Y_0$ if is a strong confounder while avoiding regression to the mean bias for treatment effect on $Y_1$?
Current License: CC BY-SA 4.0
4 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Feb 6, 2023 at 14:33 | comment | added | andrew_lor | I agree, I think your idea to include the interaction is better than just using the difference (post-pre), in this way you allow the slope to change. Moreover, I was thinking about this, using the difference (although is handy) means that you are making assumtpions on the relation between post and pre. Thank you for the reference. | |
| Feb 3, 2023 at 13:32 | comment | added | RobertF | Correct, $Y_0$ precedes treatment. We do not observe zero difference between mean $Y_0$ in trmt and ctrl groups - I'm thinking post = pre + trmt + trmt*pre would be an appropriate model. Thank you for the link. Have you read Tennant et al. "Analyses of ‘change scores’ do not estimate causal effects in observational data" (2021)? The authors argue that change scores may be useful in longitudinal RCTs but produce biased estimates in observational studies, where ANCOVA is preferred. | |
| S Feb 2, 2023 at 7:41 | review | First answers | |||
| Feb 2, 2023 at 7:47 | |||||
| S Feb 2, 2023 at 7:41 | history | answered | andrew_lor | CC BY-SA 4.0 |