Technique bias in measuring acts of altruism : The case of voluntary blood donation

This study compared personal interviews and a postal survey, posing the same questions to two random samples independently drawn from the same population. It sought reasons why respondents began and may subsequently terminate, involvement as voluntary blood donors in Toronto, Canada. The priorities of respondents (ordinal rankings) were virtually identical in the two samples, notwith-standing a clear response rate and socio-economic differential between the modes. Differences in item response rates were compatible with explanations involving interviewer bias and question threat. The study suggested the low response rates of postal surveys may not necessarily imply relevant non-respondent bias. In certain cases at least, other aspects of technique bias may be a more important research concern.