Random Digit Dialing: A Comparison to Personal Surveys

In: Public opinion quarterly: journal of the American Association for Public Opinion Research, Band 42, Heft 1, S. 105-114

Abstract

A report on the replication of a large, personal interview survey by means of a random digit dialing (RDD) telephone survey. The objective was to test the accuracy & efficiency of RDD surveys in comparison to the best traditional alternative. The personal interview survey was based on a complex probability sample of Cincinnati households, with the field work being done by the Census Bureau during Jan through Mar 1974. It was commissioned by the Law Enforcement Assistance Administration (LEAA) to collect information on crime victimizations, attitudes toward police, & perceptions of crime problems. All persons ages 12 & older (N = 19,903) from 9,708 households were interviewed. The Behavioral Sciences Laboratory at the U of Cincinnati, Ohio conducted the RDD telephone survey during Apr 1974 under a grant from the Police Foundation. The RDD questionnaire consisted of a large subset of items from the LEAA/Census instrument. It was administered to 1,639 persons in 800 households citywide, plus a supplemental sample of 1,147 persons from 662 households located in Police District 1 (a poor, rundown section of the city). The samples were compared on demographic characteristics, measures of crime victimization, & attitudes toward crime & the police. Chi-square & t-tests were used, as required, to test for sample differences. For the citywide sample, only education & perception of changes in US crime levels were significantly different at the .01 level, out of 20 tests made. The RDD sample had somewhat larger proportions of blacks & Fs. The RDD telephone survey was much less expensive per interview & permitted better supervision of the interviewers. These results indicate that RDD is an accurate & cost-effective alternative to traditional personal interviewing surveys for most research applications. 2 Tables. AA.