House Effects and the Reproducibility of Survey Measurements: A Comparison of the 1980 GSS and the 1980 American National Election Study

Abstract

This paper examines a collaborative experiment between the General Social Survey, National Opinion Research Center, and the American National Election Study, Center for Political Studies. Identical items were asked by both surveys, thereby permitting a test of reproducibility and house effects. Significant differences were found between the houses, with the largest and most systematic differences involving item nonresponse. Various explanations for the differences are examined and a general assessment of the importance of the differences is presented. Tom W. Smith is a Senior Study Director, National Opinion Research Center. This research was done for the General Social Survey project directed by James A. Davis. The project is supported by the National Science Foundation, Grant No. SOC77-03279. Public Opinion Quartedly Vol. 46 54-68 ? 1982 by The Tiustees of Columbia University Published by Elseviei Noith-Holland, Inc. 0033-362X/82/0046-54/$2.50 This content downloaded from 157.55.39.255 on Mon, 01 Aug 2016 06:02:33 UTC All use subject to http://about.jstor.org/terms