Variability of dietary assessment values among nutrition students

The effect of the interviewer's recording and evaluation of responses and calculation of nutrient data on the variability of dietary assessment results was studied. Thirty-seven dietary history interviews were examined for significant differences between the results calculated by a University Affiliated Cincinnati Center for Developmental Disorders (UACCDD) staff nutritionist and by a group of student observers. A significant difference using a paired t-test (p less than .05) for the calculated magnesium intake and the Food Frequency Summary Score for the meat group was found. The mean results for all graduate students on each variable were also compared with the undergraduate means using a group t-test. Significant differences were found in the calculation of calcium and vitamin D intake and in estimated daily caloric output. These findings agree in general with the reports of other investigators, i.e., that few significant differences of less than 10% are attributable to variability among interviewers when the mean results from large samples are examined. Since large variations between the individual results reported by different students observing the same interview were found, there is a need for extensive training and experience with interviewing, interpretation of food intake data, and familiarity with the use of a variety of food composition tables and computer data bases.