Validity of telephoned diet recalls and records for assessment of individual food intake.

Six methods of assessing individual food intake reported by telephone were compared for accuracy. A sample of 107 students eating in a dormitory dining hall was used. Two 7-day and four 3-day diet records were reported by telephone to either an interviewer or an answering device; and fourteen 6-h recalls and seven 24-h recalls were obtained by an interviewer over the telephone. To examine validity, the investigators observed respondents' intake for 28 days. Food item agreement scores comparing observed and reported data were calculated. Seven-day records were most accurate (87% food item agreement); 3-day records and 6-h recalls were equivalent (75%); and 24-h recalls were least accurate (69%). There were no significant difference in accuracy of reporting records to an interviewer versus a recording device, but respondents preferred the interviewer. Results of telephoned reporting compare favorably with personal interview techniques used by other investigators, indicating that telephoned methods should be further explored in community settings.