Abstract
In 1958 A.J.P.H. published a study by Knobloch and Pasamanick (A.J.P.H. 48: 1201-1208, 1958) linking seasonal variations to the occurrence of mental retardation. The writer has undertaken a critical analysis of that report and finds that the authors have created a correlation between season of conception and retardation principally by the improper use of statistics. Statistical tests were performed between highest and lowest values observed without regard to hypothesis. Improper statistical tests were used. The sample itself was biased and no allowance was made to correct for sampling error. The original hypothesis itself was based on hearsay and commonly held "folksy" beliefs rather than on available evidence. Finally, similar observations made on birth pattern of retarded children in Cincinnati failed to agree with the observations made by Knobloch et al. The author discusses the proper use of statistics in a study of this type and the dangers of drawing conclusions for biological questions solely by statistical manipulations.