An experimenter may "wear two hats," that of subject-matter expert and that of statistician. Is our experimenter wearing primarily the first hat, primarily the second, or both about equally when he or she:
(a) Thinks up the problem?
(b) Translates it into a statistical question?
(c) Draws a conclusion that the average performance of experimental and control groups is really different?
(d) Decides that the imposed difference in treatment was responsible for the difference in average performance?
(e) Relates this finding to those of previous studies?