---------------------------------------------------------
Roy M. Poses MD
Brown University Center for Primary Care and Prevention
Memorial Hospital of RI
111 Brewster St.
Pawtucket, RI 02860
USA
401 729-2383
fax: 401 729-2494
[log in to unmask]
----------------------------Original message----------------------------
dr birnbaum is absolutely correct and it's nice to see the same citation i
was planning on sending (it's a classic for survey work)
btw 60% is very generous, for US government work the agencies usually like
80-90% or higher [private companies may require even higher levels]
the problem at that level is the costs involved in follow-up on
non-respondents such surveys that i have conducted typically require 3 to 4
major follow-up contacts and may add telephone tracing of survey
non-respondents
a high nonn-response rate makes it very difficult to estimate non-response
bias
all of this speaks to the need for a well designed study, as rigorous as
any RCT, with well defined sampling frames, preparation of the respondents,
well targeted and well designed instruments, and diligence on the part of
the researcher
------------------------------------------------------------------------
All this is very reasonable if you are doing a survey to find out about
prevalence, e.g., a survey sent to patients about their recent dietary intake.
And I suppose this is the most common reason one does surveys.
But I would like to mention a somewhat special case. Sometimes we do true
psychological experiments in a paper and pencil format that resemble surveys.
So, for example, I might send out patient vignettes to physicians that vary
according to specific characteristics, with the goal of seeing whether the
physicians' responses vary according to these characteristics. In this case,
the response rate affects the generalizability of the study, but not the
internal validity. Thus, the minimum acceptable response rate in such a
circumstance is not so clear. Whether a less than 60% response rate
invalidates such a study is very arguable. (Although I have gotten papers
that reported such work rejected because the response rate was 60%.) Note
that in other forms of experiments, RCT's, we often accept response rates
(in terms of the proportions of patients offered enrollment who actually
enroll) of much less than 60%.
|