From: Francesca S. Alcorn (email@example.com)
Date: Wed 10 Mar 2004 - 19:43:07 GMT
>Thanks Frankie and Lawry. I must admit I'm surprised to hear that
>this is not a big area of study, although there is that 'survey
>survey' problem which is rather thorny.
I think you could find quite a bit of info out there. Try googling
"survey instrument design". Or just "survey design". "Survey design" at Amazon gives you quite a few titles.
>I found the idea of just discounting the answers of those who are
>suspected to be 'cheating' interesting -- kind of like the 'bad day
>to predict the weather' principle that comes out of complex short
>term climate models, cos that's all they can manage given the
Yeah, I thought it was pretty sneaky, probably why it stuck in my
head all these years.
>Just to put it in context; I'm predominantly a data modeller and
>user reqs gatherer these days, and getting people to (a) be
>completely honest about how they and others work, and (b) stop
>saying they're in favour of good things, and against bad things, is
>a bugger :)
Now why on earth would you want people to say that they are against
good things? :)
Is this something you can fix by using value-neutral language? Or
ask them what would be the advantages/disadvantages of "A"? Might at
least help you reframe the questions. Or (trying desperately to
bring this back to memetics) transplant the idea from one memeplex to
Can you guarantee the anonymity of the respondents?
This archive was generated by hypermail 2.1.5 : Wed 10 Mar 2004 - 19:55:29 GMT