[Training and standardization of personnel for performing an anthropometric study in school children]

In order to avoid ponderable errors among the anthropometrists working in a project of school health, a training model was developed. Four subjects were trained (A, B, C and D). Significant and highly significant differences in the reading of skinfold (F = 5.15) and height (F = 8.17) were observed only with subject C. These differences were not present in the last reading, which means that there was an improvement in the determinations. Analyzing all subjects, it was possible to verify that subject D presented significant and highly significant differences with the other subjects (skin fold delta 5% - 6.16 and delta 1% = 8.08, arm circumference delta 5% = 0.50 and delta 1% = 0.67, height delta 5% = 0.27). This subject, however, had the lowest variability in the readings, which means that this systematic error was personal. The method that was applied is useful in the training and standardization of anthropometrists, and for the detection of those subjects who present systematic errors.