Asch D A
Veterans Affairs Medical Center, Philadelphia, PA, USA.
Epidemiology. 1996 Sep;7(5):550-1.
In a survey about euthanasia, 1,600 critical care nurses were randomly assigned to receive either three complete, anonymous mailings of the questionnaire or, with each mailing, a coded postcard to be returned separately from the questionnaire to reduce subsequent mailings to previous responders. The response rate in these two groups was 76.5% [95% confidence interval (CI) = 73.6-79.4%] and 69% (95% CI = 65.7-72.4%), respectively. The two strategies yielded similar responses, and costs were much lower for the postcard group. Using coded postcards to be returned separately from completed instruments appears to lower the response rate to anonymous mail surveys, but it also lowers cost and may not introduce additional bias.
在一项关于安乐死的调查中，1600名重症监护护士被随机分配，要么收到三份完整的匿名问卷邮件，要么每次收到一份编码明信片，明信片要与问卷分开返还，以减少后续对之前回复者的邮件发送。这两组的回复率分别为76.5%[95%置信区间（CI）=73.6 - 79.4%]和69%（95%CI = 65.7 - 72.4%）。这两种策略产生了相似的回复，并且明信片组的成本要低得多。使用与填好的问卷分开返还的编码明信片似乎会降低匿名邮件调查的回复率，但它也降低了成本，并且可能不会引入额外的偏差。
