lmer Error: number of observations < number of random effects

问题

I am trying to run an LMM allowing time to be a random effect:

lmer(efficacy ~ time + (1+time|id), data=long, REML=TRUE)

However, I get the error messageas below:

Error: number of observations (=3948) <= number of random effects (=4496) for term (1 + time | id); the random-effects parameters and the residual variance (or scale parameter) are probably unidentifiable

I have three time points but not all subjects have data for all 3 time points (attrition) but my missing count on my outcome variable (efficacy) only had 13 missing values and I have a total number of 3961 participants in my entire data set. If anyone can shed some light on this or point me in the right direction I'd much appreciate it.

Thanks.

来源：https://stackoverflow.com/questions/53699081/lmer-error-number-of-observations-number-of-random-effects