The effects of two training methods on the accuracy of skinfold measures



Journal Title

Journal ISSN

Volume Title



The purpose of this research was to evaluate the effects of two training methods on the accuracy of skinfold measures. Six physical education teachers, inexperienced in measuring skinfolds, were randomly assigned to two groups of three testers. One group was trained by reading written instructions; the other group viewed a video tape on the skinfold measurement technique. Lange calipers were used to measure ninety-five college students at the tricep and subscapular sites. Measures by two experienced testers taken on forty-nine of the subjects served as a criterion score. Median reliability coefficients for groups of testers were .84 (readers), .94 (video tape) and .92 (experienced testers). Reliability estimates predicted using the Spearman Brown Prophecy Formula were .63 (readers), .84 (video tape) and .92 (experienced testers) for a single tester. Generalizability analyses enabled quantification of sources of variance in the model. The percent of variation for subjects ranged from 60.46% to 86.95% depending upon the site assessed. Zero percent of variation existed for methods when comparing the training groups. This increased substantially when comparing the novices to the experienced testers. The experienced testers obtained higher values for all measures. It is concluded that either training method elicits reliable skinfold measures, but the validity of the scores is questionable.



Skinfold thickness, Body composition