What is the difference between
- generalized linear mixed models, and
- linear mixed effect models (lmer function in package lme4)
in terms of distributions of the response variable? Do they both work with non gaussian distributions?
$\begingroup$
$\endgroup$
2
-
2$\begingroup$ It'd probably be more useful to read up on them a little yourself - starting with e.g. en.wikipedia.org/wiki/Generalized_linear_mixed_model - & ask focussed questions here to resolve any doubts you might have. $\endgroup$– Scortchi ♦Commented Nov 18, 2015 at 14:05
-
$\begingroup$ Thanks for the link :) I've read it and updated my question...my doubt is regarding the assumption that lmer does...I thought it was relaxing the assumption of having a gaussian distribution for the response variable. $\endgroup$– gabboshowCommented Nov 18, 2015 at 15:02
Add a comment
|
2 Answers
$\begingroup$
$\endgroup$
1
Linear mixed-effects models describe the relationship between a response variable and independent variables, with coefficients that can vary with respect to one or more grouping variables. A mixed-effects model consists of two parts, fixed effects and random effects. Fixed-effects terms are usually the conventional linear regression part, and the random effects are associated with individual experimental units drawn at random from a population. The random effects have prior distributions whereas fixed effects do not.
The Generalised Linear Mixed Model as linear predictor contains random effects in addition to the usual fixed effects, but would be estimated as a one step regression rather than Expectation Maximisation model.
-
$\begingroup$ Hi Thanks for your answer, but also linear mixed effect models contains random effect...isn't it? $\endgroup$ Commented Nov 18, 2015 at 16:45
$\begingroup$
$\endgroup$
You might be mixing up general linear models and generalized linear models.
Linear mixed models assume your response (or dependent) variable is normally distributed. Generalized linear mixed models do not; instead you have to provide a suitable distribution and link function for your data.