Atkinson & Batterham 2015, True and false interindividual differences in training response

Single pre-post designs inflate the responder/non-responder spread; a control group is needed to separate true response from noise. Strong.