[Skip to Content]
[Skip to Content Landing]
November 22, 1985

Observations vs Statistics-Reply

Author Affiliations

American Medical Association

JAMA. 1985;254(20):2896. doi:10.1001/jama.1985.03360200046026

In Reply.—  Regrettably, Dr Feldman has missed the major point of my editorial, namely, that "final recommendations concerning the suitability of manuscripts for publication are made in consideration of the overall quality of the studies submitted, the careful documentation of results and interpretation of findings, and the perceived validity and importance of research conclusions."1 Within the peer review process, the statistical reviewer may be called on to ascertain whether reported results are based on cogent observations and logical reasoning, as well as on appropriate data analysis. Therefore, the concept of statistically sound experiments that are flawed by unsound logic is incongruous and erroneous. It is the very preservation of basic principles of sound reasoning that constitutes the primary concern of the statistical reviewer—more so than the degree of statistical sophistication exhibited in any manuscript.Although some of the studies cited by Dr Feldman may not have benefited from statistical