Hi experts,
I am trying to examine "homophily" in school students. That is, I am examining whether pairwise similarity in various traits correlates with closeness in social networks.
Each row of my spreadsheet contains data for following columns: child_1_id, child_2_id, social closeness of child_1_id and child_2_id, pairwise similarity in trait of interest of child_1_id and child_2_id. This repeats for all pairwise combinations of children. The simplest form of my model is therefore the following:
trait similarity ~ social closeness + (1 | child_1_id) + (1 | child_2_id)
However, I would like to nest within classrooms. Importantly, trait similarity and social closeness were only calculated for children in the same classroom, and therefore child_1 and child_2 are always from the same classroom.
How do I go about nesting the crossed random effects for child id within a term for classroom?
Thanks in advance,
Luke
See these links:
https://stats.stackexchange.com/questions/228800/crossed-vs-nested-random-effects-how-do-they-differ-and-how-are-they-specified
https://www.muscardinus.be/2017/07/lme4-random-effects/
http://lme4.r-forge.r-project.org/book/Ch2.pdf
https://psyteachr.github.io/ug3-stats/linear-mixed-effects-models-with-crossed-random-factors.html
I think what you are aiming for (correct me if I am wrong) is to detect effects on the child level (social closeness), on the pair level (similarity in trait of interest) and on the classroom level (mainly as rando effects). In your current dataset, I think it ill be difficult to tell lme4 to investigate just that. I guess to do the intended analyses you have to restructure your data so that you have each child in a row with the columns Child_ID, Pair_ID, Classroom_ID, Social_Closesnes, and Similarity.
The model structure would look something like this:
trait similarity ~ social closeness + (1 | Child_ID) + (1 | Pair_ID) + (1|Classroom_ID)
P.S. as every Child is only part in one pair and one classroom, I'd say that this is not a crossed but rather a nested structure, but maybe I am just picky.
Best
Sven
May this link answring your question
https://www.muscardinus.be/2017/07/lme4-random-effects/
- Badges
- Science topic
- Similar topics
- Medicine
- Epi
- Age Groups
- Children