2

What does this formula mean?

lmer(y ~ (1|g1/g2))

equivalently:

lmer(y ~ (1|g1) + (1|g1:g2)

According to the PDF that I found it in, linked here, it means:

"Intercept varying among g1 and g2 within g1"

I would love it if someone would elaborate on this, and would differentiate it from, say

lmer(y ~ (0 + g2|g1))

Thanks!

goldisfine
  • 626
  • 7
  • 16
  • Isn't the answer available at http://stats.stackexchange.com/questions/13166/rs-lmer-cheat-sheet/13173#13173? – whuber Mar 17 '15 at 23:28
  • @whuber, I'm a big fan of that answer, but I wonder if there is some room for someone to elaborate specifically on the nesting syntax `/` & how that relates to the more common syntax. – gung - Reinstate Monica Mar 17 '15 at 23:31
  • @whuber I really did see that post before (have the tab open from the morning). Mike Lawrence does an awesome job in that answer and it cleared up a lot of things for me. However, no where in that answer is there syntax like (1|g1/g2) or (1|g1:g2) and I believe, based on the syntax in the pdf linked in the question, that this means something different. – goldisfine Mar 18 '15 at 00:30

0 Answers0