Mixed mode Data Collection: Telephone and Personal Interviewing

Examined the utility of a mixed-mode telephone interview/personal interview method of data collection in a study of voting in union representation elections. The sample of 1,239 employees in 31 different elections was heterogeneous with respect to age, education level, sex, race, wage rate, and urban-rural background. Ss who could not be contacted by telephone or who refused to participate in a telephone interview were interviewed in person. The utility of the mixed-mode method was evaluated with respect to (a) response rate (92%), (b) cost (saved $10,746 in interviewer wages alone), and (c) quality. On most indices, the quality of the data collected in person and by telephone was similar. Respondents interviewed by telephone were more likely than those interviewed in person to refuse to disclose their vote and less likely to report unlawful campaign practices. Implications for the internal validity and generalizability of the voting study are discussed.