MULTIPLE CORRELATIONS AND BONFERRONIS CORRECTION

Correlation coefficients between biological measurements and clinical scales are often calculated in psychiatric research. Calculating numer ous correlations increases the risk of a type I error, i.e., to errone ously conclude the presence of a significant correlation. To avoid thi s, the level of statistical significance of correlation coefficients s hould be adjusted. Threshold levels of significance for correlation co efficients were adjusted for multiple comparisons in a set of k correl ation coefficients (k = 1, 5, 10, 20, 50, 100) by Bonferroni's correct ion. Significant correlation coefficients were then calculated accordi ng to sample size. The change in the threshold values of significance is larger when the number of correlations goes from 1 to 5 than when i t goes from 50 to 100. A correlation coefficient, statistically signif icant at 5% when calculated alone, can be under the threshold level of significance when calculated even among a few other coefficients. Foc using on the most relevant variables or the use of multivariate statis tics is advocated. Biol Psychiatry 1998;44:775-777 (C) 1998 Society of Biological Psychiatry.