Timeline for Mixed effects modeling using R with time varying predictors
Current License: CC BY-SA 4.0
4 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 4, 2020 at 18:49 | comment | added | EdM | @student_R123 the paper dealt with up to 4 measurements over time, so a first-stage modeling of phthalate values over time was needed to get slopes and intercepts to test, in the second-stage model, whether changes over time were associated with outcome. With only 2 measurements you can go directly to a single model incorporating the continuous predictor, say, as (1) its mean value and (2) its difference. For the binary categorical (say, T/F), try a single 4-level categorical: both T, both F, T->F, F->T. Not sure that a random-effect model, with its added assumptions, would help more. | |
| Apr 4, 2020 at 18:12 | comment | added | student_R123 | In this paper they have discussed a two stage model. But they mentioned that it will not be suitable when the predictors are categorical. | |
| Apr 4, 2020 at 18:10 | comment | added | student_R123 | Thank you for the answer. I came up to the conclusion that mixed effects model is appropriate, based on this paper: ncbi.nlm.nih.gov/pmc/articles/PMC4417225 | |
| Apr 4, 2020 at 17:48 | history | answered | EdM | CC BY-SA 4.0 |