Optimal Times to Contact Sample Households

Abstract

Finding a suitable respondent at home is an essential and expensive component of a household survey. This article reports on the results of a study of the probabilities of finding someone aged 14 or older at home and discusses the application of such data to survey design and budgeting. The authors are staff members of the Research Triangle Institute, P.O. Box 12194, Research Triangle Park, North Carolina 27709. M. F. Weeks is a survey specialist in the Survey Operations Center; B. L. Jones, R. E. Folsom, Jr., and C. H. Benrud are statisticians in the Sampling Research and Design Center. Public Opinion Quarterly ? 1980 by The Trustees of Columbia University Published by Elsevier North Holland, Inc. 0033-362X/80/0044-101/$1.75 This content downloaded from 157.55.39.129 on Thu, 28 Jul 2016 04:19:54 UTC All use subject to http://about.jstor.org/terms 102 WEEKS, JONES, FOLSOM AND BENRUD In 1972 the U.S. Bureau of the Census released a report showing the probabilities of finding people of various characteristics (sex, race, and age group) at home at various times of day on a weekday during November 1971 (Weber and Burt, 1972). This report also included a limited comparison with 1960 census data, which showed a significant decrease in the proportion of households with at least one person aged 14+ at home. The census report, however, included no data for specific days of the week or for weekends. During the period from March to June 1976, the Research Triangle Institute (RTI) screened 22,290 occupied housing units in its national area probability sample to obtain household rosters from which to identify persons eligible for a household survey being conducted concurrently as well as for subsequent surveys involving different target populations. Among the data recorded for each screening visit were the time of day, day of week, whether or not a household respondent aged 14+ was at home, and, if so, the race of the respondent. However, no data were obtained on who was at home, other than the respondent, at the time of the successful screening visit. This article reports on an analysis of these screening data to estimate the proportion of households with at least one person aged 14 or older at home during spring 1976. Comparisons with census data are made to indicate trends in the proportion of such households, by time of day and type of area. Also shown are estimated proportions for each day of the week (including weekends) and approximate estimates of standard errors for these proportions. The following sections provide additional detail on the RTI survey, describe the format and content of the tables and figure in the article, point out some of the highlights from these data, and comment on the utility of the data to the survey researcher in designing and budgeting a household survey.