A set of training materials for professionals working in intervention epidemiology, public health microbiology and infection control and hospital hygiene.
Need help with your investigation or report writing? Ask the Expert. Free advice from the professional community.
You can't make decissions on this page's approval status because you have not the owner or an admin on this page's Group.
Jean Claude Desenclos
As mentioned before, it is very important to pretest ("pilot") the questionnaire. The questionnaire should be piloted with a similar group of people to your intended subjects. The aim of a pilot survey is to obtain estimates about the expected response rates, data quality, the validity and comprehensibility of the questionnaire (1).
It allows highlighting problems such as inappropriate questions or ambiguity before starting the real survey. The effects of an alternative wording should be tested as well. A pilot questionnaire should also include questions about the overall impression on respondents and interviewers in order to include their comments. It might be useful as well to ask non-respondents for the reasons for not participating.
Often the pilot study leads to several amendments before the survey starts. Therefore, there should be sufficient time allowed for this phase in the orginal time schedule ofthe study.
1. Silman AJ and Macfarlane GJ: Epidemiological studies. 2nd edition 2001, Cambridge. Pp 138-141
Join the discussion about this article in the forum!
Arnold Bosman posted on 7/5/2011 9:01:11 AM:
We could debate whether or not we are able to 'measure' a causal effect. In my view, what we measure are observations (counts) expressed in numbers, rates, risks for example.
The comparison is already a computation, resulting in an 'effect', which caj be consided an estmation of the effect in the population.
If the effect (e.g. measure of association) is causal or not, cannot be measured, not tested. It can merely be inferred.
marcelius atanga replied on 6/16/2015 10:01:13 AM:
I think even when we take a survey and count, at the end the numbers obtained are only estimates,
and so more important to me is to what extend should an estimate be consider useful, in fact what are tolerable error margins, given that no 2 situations are truely comparable?
Arnold Bosman replied on 6/16/2015 10:33:11 AM:
Perhaps we should open a philosophy wiki?
You need to be logged in to post comments.
You can log in here. You can register here if you haven't done so yet.