Goyal and his colleagues graded the

Goyal and his colleagues graded the 47 remaining studies on strengths of evidence — how confident could they be that the results they were reading were real and sufficiently unbiased? A high strength of evidence meant the data were consistent — good-quality studies, good measures of the outcome, his team could be fairly sure that a certain effect really had or had not been produced, and that further studies could replicate the data and add to what we know. After “high” strength of evidence came moderate, then low, then insufficient.