Abstract

My reflections last issue on experimental replicability prompted some further thoughts on the subject. In particular, I wondered to what extent we should consider scientific expertise to be an art, or something more like a gift than a skill. In that previous Editorial I criticized a familiar view expressed as follows by Karl Popper: “Any empirical scientific statement can be presented (by describing experimental arrangement, etc.) in such a way that anyone who has learned the relevant techniques can test it” (Popper 1959:99, emphasis added). I noted that given the inevitable differences between original experiments and replication attempts—magnified in the behavioral sciences by many additional kinds of potentially relevant variables (such as well-documented experimenter effects), it may be unreasonable to expect success when replication attempts are conducted by someone other than the original experimenter. What I want to consider more closely now are the related questions: What are the relevant techniques? Can they be captured and conveyed by a mere list of procedures, like a recipe for baking bread? To what extent can these techniques even be learned?