...small-sample, well-designed studies may have sufficient statistical power to produce scientifically valid results, and failure to disprove a null hypothesis may sometimes challenge a well-established body of evidence...raw data must be made available..