Dear research community!

I am currently conducting a meta-analysis for my MSc thesis. My funnel plot is asymmetrical and indicates heterogeneity between studies rather than publication bias.

After I excluded three studies whose methodology was more heterogeneous than that of the other studies, my funnel plot became symmetrical. I am now wondering: when I report my results, should I report the results that INCLUDE or EXCLUDE those three studies? I am thinking I should report those results that EXCLUDE the three studies, as a meta-analysis aims to synthesise studies that have a very similar methodology.

On the other hand, those three studies did not use a strikingly different methodology. Basically, I am looking at the amount of 5-HT that is released after inflammatory challenges in rodents. Those challenges are mainly cytokines or LPS administration, but some studies have used carrageenan and formalin, and one study used capsaicin. Thus, heterogeneity stems from studies using different inflammatory stimuli and different doses. So, I am currently not sure if that is heterogeneous "enough" for these studies to be excluded.

Thus, I am wondering: which results should I present - the ones that include these three "outlier" studies or the ones that exclude them?

More Lucienne Billy's questions See All
Similar questions and discussions