Excerpt 11.24 A Power Analysis Used to Determine Sample Sizes

The total number of clinical dietitians in the district directories was 412. From this population, a random sample of 205 dietitians was selected. The same number of internal medicine physicians was randomly selected from the physician directory. The sample size of 205 per occupational group was determined according to power tables of Kraemer and Theiman. Given an approximate return rate of 50% per group (i.e., 103 per group), an anticipated small effect size of .24, and using a two-tailed t-test with a set at the .05 level, the sample size results in 80% power, which is the recommended level. Source: M. E. Boxhtari and B. J. Cardinal. (1997). The role of clinical dietitians as perceived by dietitians and physicians. Journal of the American Dietetic Association,97(8), p. 851.

Excerpt 10.20 Use of the Bonferroni Technique

Independent variables were correlated with liaison satisfaction. Because 38 correlations were conducted, an a/N correction was used to reduce the risk of Type I error. To be significant, a correlation between a variable and liaison satisfaction required a p-value of .001 or less. Source: L. Bennett and S. Coe. (1998). Social work field instructor satisfaction with faculty field liaisons. Journal of Socail Work Education,34(3), p. 374.