Personal versus telephone surveys for collecting household health data at the local level.

Personal and telephone interview surveys were conducted simultaneously during 1981 in the same area (four counties in the area of Tampa Bay, Florida) and utilizing the same interview schedule. Following completion of the surveys, validity checks were made with the medical providers reported by a subsample of respondents to each mode. The telephone survey yielded a lower response rate but cost less than half the personal interview. There was some evidence of nonresponse bias in the telephone survey, and some relatively minor differences in responses were found between the two modes, but there was no conclusive evidence that the response differences resulted from mode effects. Telephone respondents appeared to be somewhat more accurate in their reporting of visits to medical providers, although accuracy comparisons must be interpreted with caution in view of the disparate success experienced for the two modes in securing permission forms for the release of medical record information.