[R-meta] MLMA - shared control group

[Viechtbauer, Wolfgang (SP)]

>-----Original Message-----
>From: Reza Norouzian [mailto:rnorouzian using gmail.com]
>Sent: Monday, 30 August, 2021 17:36
>To: Viechtbauer, Wolfgang (SP)
>Cc: Jorge Teixeira; R meta
>Subject: Re: [R-meta] MLMA - shared control group
>
>Dear Wolfgang,
>
>Jorge may benefit from using cluster-robust estimates of its fixed
>effects in his (perhaps 3-level) model. However, my current
>understanding is that even assuming Cov(e_{ijk}, e_{ij'k}) = 0 for two
>observed estimates from a person on say two outcomes in the same study
>while that is, in reality, not the case (perhaps in a major way),
>gives Type I error rates and confidence intervals' coverage that are
>nearly accurate.

Maybe or maybe not. You are essentially hoping that the bias in the variance components in this misspecified model compensates for the misspecification. That's does not sound like an approach I generally would rely on.

Best,
Wolfgang