I watched the following video yesterday and realised how educational it is. One area that sticks is regarding meta-analysis. It allows say 10 bad studies with negligible positive association, to show an improved association cumulatively (naughty to say the least). The statistical methodology used also makes a difference to outcomes. I knew shenanigans were employed to obtain desired results, but after this deeper dive my cynicism for studies has increased.
He seemed to miss the obvious that doing a meta analysis of 10 really bad studies doesn't in fact make the meta analysis any better. GIGA is still garbage?