Citation

Smith TW. Soc. Sci. Res. 2002; 31(4): 576-593.

Copyright

(Copyright © 2002, Elsevier Publishing)

DOI

10.1016/S0049-089X(02)00015-7

PMID

unavailable

Abstract

Different methods of measuring inter-racial contacts produce large differences in the level of the same. Experiments conducted on the 1998 General Social Survey indicate that estimates of how many people have inter-racial friends are much higher when a direct question is asked, intermediate when a three-step procedure is used, and the least when a name-generator, network approach is used. Cognitive and social desirability reasons for these effects are discussed.