In developmental science, we typically test small samples of a very noisy population, so both meta-analyses and replications seem like a logical answer to the question, how robust results really are. But what can we learn from each and how do both fare when combined?
Christina Bergmann (Thu,) studied this question.