Using computer kiosks can be quite a labor-saving  strategy because the
visitor enters the data directly instead of a data entry person entering
information from a survey or an interview form.  Here are a few hints:

1.      To prevent bias you really need to have someone stationed at the
kiosk inviting visitors in a random sequence to take part in the survey
and noting who refuses.  You then compare the sample you have with your
overall visitor profile for that season using key characteristics, such
as visiting with children/no children; visiting first time/repeat
visitor; local person or tourist.  By weighting the sample properly you
correct for some of the bias.
2.      You can do the same in comparing your pre- with your post-,
again to make sure you're comparing apples with apples.
3.      You'll probably want to test out your questions using
traditional interview methods and including some open-ended prompts to
your questions.  ("You chose "b", why?" or "You said you liked the
program...can you give me some specifics about that?") before loading
into the kiosk That way you have a better idea what the visitor means by
some of the responses.  It's harder to do open-ended questions in a
kiosk than in an interview.

I'd sure like to hear more about the project when it's done

Linda Wilson
Visitor Studies and Evaluation
Shedd Aquarium
Chicago, IL 60605
[log in to unmask]