Much interest and attention have recently been focused on census underenumeration, particularly undercounts of poor urban minority populations (Hear 1968, Trans-Action 1968). As an outgrowth of this concern, the work reported here is part of a joint study by the authors and the Center for Research in Measurement Methods. Our purpose is to compare the results of interview-survey research with data produced by the participant-observational method of ethnography and cultural anthropology. We believe this comparison will be revealing with respect to the usefulness of questionnaire interview techniques, the accuracy of basic social statistics on poor minority urbanites, and the validity of both theories and policies based on these statistics.

The central procedure of this research has been a double-blind comparison of data collected by the two methods from the same sample population. Since mid-1968 the two authors have been carrying out a resident ethnographic study of the predominantly Afro-American inner-city community which we refer to as Blackston (Valentine and Valentine 1970a, 1970b, 1971a). During the summer of 1969 the Research Center sent CPS, HIS, and QHS survey interviewers to 33 dwelling units in a section of this community which we have been studying intensively. The interviewers were unaware of our work. The Center identified for us the dwelling units surveyed, but we received no data from the interviews. Before, during, and since the period of interviewing we have been observing the households covered in the surveys, along with a much larger number of other households. In 1970 we completed a preliminary report for the Center detailing the ethnographic finding, as of the time when the interviews were carried out (Valentine and Valentine 1970c). We have since received the data tabulated by the interviewers, which makes possible several comparisons.