Is there any advantage of a paired t-test over a simple mixed effect model?

[eric some]

Dear All,
I aim to assess the change (difference) in mothers'body mass index every month over the course of breastfeeding. In the non adjusted analysis I used a paired t-test. I wanted to ask if there is any advantage of using repeatedly a paired t-test over a simple mixed effect model or vice versa. In other terms, which method will be more appropriate for my data.?

[Marcos Almeida]

There is no advantage in using "repeatedly" a paired t-test. In fact, you should "pay a high price" for that, in terms of adjusting for familywise error. In short, as you know, the mixed model will allow to "control" for covariates, withing group variation and between group variation.

You asked "which method will be more appropriate" to your data. However, you didn't inform much. For example, sample size, predictors, how many repeated measures, etc.

That said, the first paragraph still applies to "complex" data analysis, something expected in studies with human beings.

[Carlo Lazzaro]

Eric:
Example 1 under -mixed- entry, Stata .pdf manual can be helpful.