The use of the multiplicity survey to identify migrants.

New approaches must be developed and tested to meet the needs for current data on population movement. One such approach is the survey with multiplicity (social networks). It allows respondents to report events (in this case, migration) that occurred to specified relatives living elsewhere. In order to test the value of the approach on a limited, exploratory basis, a two-stage research design was employed. A sample of persons resident in Cranston, Rhode Island who had changed address in the three years preceding the survey, was first identified and information was gathered in Round I on their mobility and on their relatives living in the State. In turn, these relatives (Round II) were asked about the mobility of their parents, siblings, and children. (The responses from Round II should have included the migrants in Round I.) Evaluation of the accuracy of Round II responses in comparison to those obtained in Round I was undertaken in terms of kin relationships, recency of move, and type of move. The results indicate that respondents had a low awareness of their relatives' movements, that moves of females were more likely to be reported than those of males, that recent moves were more frequently cited and that moves between communities were more likely to be reported than intracity mobility.