What is the difference between lmer(.. (1|groupingfactor1) +(1|1|groupingfactor2)) and lmer(..(groupingfactor1|groupingfactor2)?

Asked Jul 08 '13 at 12:36

Active Jul 08 '13 at 12:57

Viewed 624 times

I am quite new to lmer and quite confused about how to select the random effects part of the model.

The data I am working on is divided into dependent and independent variables. Each variable has two grouping factors. 1). sampling time (sample 1:7) and sampling column (1:24, where 1:3, 4:6 and so are replicates).

I have explored these different possibilities, but I am not sure on which is the correct one to use.

fm1 <- lmer(dependent ~ independent +(1 + independent|Sample)+(1+ independent|Column), data=Data)

fm2 <- lmer(dependent ~ independent +(1|Sample)+(1|Column), data=Data)

fm3 <- lmer(dependent ~ independent + (Sample|Column), data=Data)

fm3 is the model with the lowest AIC and BIC values, but I am not sure that this is the right one to use?

edited Jul 08 '13 at 12:57

Nick Cox

48,377
8
110
156

asked Jul 08 '13 at 12:36

Jacob Bruun

3

Hi Jacob and welcome! [This post](http://stats.stackexchange.com/a/13173/21054) provides some excellent explanations of `lmer`s formulas and their interpretation. What the "right" model is depends greatly on your research question and if it makes sense given your data. – COOLSerdash Jul 08 '13 at 12:46

What is the difference between lmer(.. (1|groupingfactor1) +(1|1|groupingfactor2)) and lmer(..(groupingfactor1|groupingfactor2)?

0 Answers0