Wat is een response van 80% waard?: een onderzoeksnotitie

Samenvatting

In an anonymous survey by mailed questionnaires, the return proportion was 80% of all farmers in a Dutch region. It was tempting not to worry about representativeness with regard to the variable 'visit by a change-agent.' Yet, one must assume that the real difference between respondents & non-respondents is small. A further investigation of a sample of farmers showed a significant difference between types of respondents to the anonymous surveys Since assumptions about differences between respondents & non-respondents may be false, a high proportion of returns does not need to imply representativeness. 2 Tables. HA.