Estimating Statewide Health-Risk Behavior: A Comparison of Telephone and Key Informant Survey Approaches

A statewide key informant survey of health-risk behavior was implemented in New Mexico at the same time that a random-digit-dialed telephone survey was conducted. The results obtained from the area of the state that was sufficiently sampled by both surveys demonstrated similarities in the responses on a majority of the assessment items. Differences in responses are explained by a tendency of those directly interviewed to overreport "healthy"behavior and underreport "unhealthy"behavior. The key informant approach was also able to capture information about native and rural populations not represented in the telephone survey. A reliability test of the key informant survey was conducted, and the data indicate a significant overall correlation between the two sets of responses.