cf R package geepack, contains things about longitudinal data

https://www.jstatsoft.org/index.php/jss/article/view/v015i02/v15i02.pdf

Without seeing the article, my guess is that they had repeated measures within patients (or some other unit), so they used GEE to model to correlated nature of the data.  I suspect that their use of the term, "random effects", refers to the repeated measurements wiithin patients or units.  GEE does not model random effects, rather considers the clusters or units as nuisance parameters, used only to account for the lack of independence among observations.  Their approach to the data analysis, however, may be satisfactory -- only their description confusing.

Paul

Thanks Jean

I have seen this sort of misstatement before, when indeed the data analyst had used a random effects model but the person writing the abstract was confused.  I would suggest taking a close look at the methods section to see if the modeling approach is specified there.

Thanks John for your input, what I found was they used General Estimating Equations (GEE), which cannot be used with a random effects model, so they have misspecified, along with a number of other errors throughout the paper. Thanks again, I am glad to know others have seen this type of misstatement before!

Conclusions are different  if one is using random effect models. The main conclusion will be about the variances, rather than the means. Secondly, the test statistic used are basically depend on the assumptions about the effects, i.e, whether they are purely random, purely fixed or mixed.