Timeline for Comparing "Age of emergence" or "Age of onset" in a statistical model
Current License: CC BY-SA 4.0
5 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Mar 23, 2021 at 18:07 | vote | accept | Questioner4857 | ||
| Mar 23, 2021 at 18:07 | comment | added | Questioner4857 | Excellent, thank you, I'm really grateful for your explanations! | |
| Mar 23, 2021 at 17:14 | comment | added | EdM | @Questioner4857 the choice is up to you, based on what hypothesis you wish to test. If the two Behaviors can start in either order with neither one thought to affect the initiation of the other, then (absent censoring) a simple correlation might be OK. Regression involves treating one as a predictor and the other as an outcome, which might not be appropriate depending on your situation. Survival modeling provides potentially richer analysis of timings and multi-state situations, but that might not be required to test the specific hypothesis you have in mind. Choose based on your needs. | |
| Mar 23, 2021 at 17:02 | comment | added | Questioner4857 | Thank you for your comment, I appreciate it. Your first suggestion is a nice solution that I'd not thought of. And I was not aware of multi-state models, so thank you for mentioning those. To clarify- are you suggesting that the kind of data I'm talking about should only be analysed using some kind of survival modelling? I.e., a linear regression/correlation should be avoided? I ask because really a simple association is what I'm primarily interested in. And I'm also wondering what the advantages of survival modelling are beyond censoring (in reference to the start of your comment). Thanks! | |
| Mar 23, 2021 at 15:40 | history | answered | EdM | CC BY-SA 4.0 |