Dear friends. I am conducting a meta-analysis on the effect of childhood trauma on suicide ideation and suicide attempts. The childhood trauma included 5 facets (or subscales in the most common questionnaire). I would like to run a meta-regression in which I can include all 5 subscales as predictors and have the suicide measures as the outcome (in separate models). My goal is to be able to get the relative association of each subscale with the outcome while ?accounting? for the other trauma facets, similar to how a multiple regression works. I found that a <>model-based meta-analysis <> could be the way to go, but I have no experience with these models myself. Does anyone has experience fitting model-based meta-analysis and know if this would be a good method, or does anyone has other suggestions for how I can test this hypothesis in a meta-analytic framework? Thank you for your help, Eran
[R-meta] meta-analysis (regression) with multiple predictors
2 messages · Eran Barzilai, Michael Dewey
Dear Eran What you describe is a classic meta-regression. You do not say which package you are proposing to use so I suggest you download either meta or metafor from CRAN and investigate their facilities. For what you need there is no practical difference between them. Then come back with any specific queries about the commands you are proposing to use. Michael
On 22/03/2022 01:28, Eran Barzilai wrote:
Dear friends. I am conducting a meta-analysis on the effect of childhood trauma on suicide ideation and suicide attempts. The childhood trauma included 5 facets (or subscales in the most common questionnaire). I would like to run a meta-regression in which I can include all 5 subscales as predictors and have the suicide measures as the outcome (in separate models). My goal is to be able to get the relative association of each subscale with the outcome while ?accounting? for the other trauma facets, similar to how a multiple regression works. I found that a <>model-based meta-analysis <> could be the way to go, but I have no experience with these models myself. Does anyone has experience fitting model-based meta-analysis and know if this would be a good method, or does anyone has other suggestions for how I can test this hypothesis in a meta-analytic framework? Thank you for your help, Eran [[alternative HTML version deleted]]
_______________________________________________ R-sig-meta-analysis mailing list @ R-sig-meta-analysis at r-project.org To manage your subscription to this mailing list, go to: https://stat.ethz.ch/mailman/listinfo/r-sig-meta-analysis