A bigger danger is publication bias. collect 10 well run trials without knowing that 20 similar well run ones exist but weren’t published because their findings weren’t convenient and your meta-analysis ends up distorted from the outset.
A bigger danger is publication bias. collect 10 well run trials without knowing that 20 similar well run ones exist but weren’t published because their findings weren’t convenient and your meta-analysis ends up distorted from the outset.