Hi,

We are working on a meta-analysis of VBM studies that have compared violent individuals to controls. We would like to conduct meta-regression analyses to examine the influence of the level of psychopathic traits on the results. Here are our two questions:

1. These VBM studies have used different measures/questionnaires/interview to assess psychopathy in their participants. Is this a problem and how should we deal with this?

2. Do we actually need to obtain from the authors the raw values/data of the psychopathy measures to run the meta-regression analysis?

Thank you in advance for your help!

Regards,

Stephane

Dear Stephane,

regarding question 1, I would say that this is a potential issue in all meta-analyses, not related indeed to neuroimaging. I think that you should find some way to convert the different measures to the same scale, though unfortunately this may be difficult or even not possible sometimes.

regarding question 2, you only need the mean psychopathology score of each sample, i.e. one number per study.

Best wishes,

Joaquim