Estimating the proportion of homes with functioning smoke alarms: A comparison of telephone survey and household survey results

Administering questionnaires by telephone has become an increasingly acceptable and widely used means of doing health research.(1) Telephone surveys have many advantages over in-person surveys. They are usually less expensive,(1-5) and they minimize sampling error when the sample is drawn from random-digit dialing.(1,6,7) Telephone surveys allow for timely data collection for states and subgroups of the population.(2-4,8) The main disadvantage of telephone surveys has been the potential for sampling bias(2,9) and variable estimates of behavioral factors caused by excluding households not accessible by telephone. According to the US census, 95% of US households and 91% of Oklahoma households had telephones in 1990. To determine the proportion of households with functioning smoke alarms in a low-income area of Oklahoma City with a high rate of residential fire-related injuries,(10) we conducted a targeted random telephone survey; to validate the telephone survey, we conducted a retrospective random household survey. The purpose of the survey was to establish a baseline rate for use in evaluating a large smoke-alarm giveaway program in this high-risk area.