Timeline for How to handle Pseudoreplication - Aggregate observations per group for linear mixed model or take individual observations
Current License: CC BY-SA 4.0
4 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 15, 2023 at 12:22 | comment | added | Robert Long | Hi Julia you are very welcome. Yes, what you just wrote makes sense. It would be an upper level (Site level) variable. Thanks for accepting my answer. | |
| Nov 15, 2023 at 12:15 | vote | accept | Julia | ||
| Nov 15, 2023 at 12:15 | comment | added | Julia | Thanks a lot for this answer - much appreciated! I didn't realise that I could simply nest treatment in site, but thought I had to recode observation. Just one follow-up question. To get an understanding about possible explanatory variables for the treatment differences between sites, I want to include further effects in the model (mostly environmental variables, like e.g. precipitation). My model would then look something like: lmer (B~Treatment+Precipitation+(1|Site/Treatment). Precipitation is just given once per site. So do I then just repeat the same Precipitation value within each row? | |
| Nov 9, 2023 at 14:13 | history | answered | Robert Long | CC BY-SA 4.0 |