I would suggest using a line-up. I have done a photo line-up with students and they
are fairly accurate when the target is present in the line-up. However, you should do
a target absent condition also. This condition is where they make a lot of false
alarms. Make sure that the students and dogs don't know the people who are part of
the line-up. For dogs they would do a scent line-up. Of course the similarity of the
people in the line-up contributes to the level of difficulty in identifying the
target. For the dogs, the problem might be in judging the similarity of people in
terms of smell. I can think of all sorts of fun manipulations that one could do in
addition to the basic test.
See if the dogs can differentiate between identical twins.
See if strong odorants applied after the incident can interfere with detection.
See if eating lots of garlic would interfere with detection.
I think Stephen Black is right that it would not work to just have the dog enter the
room and try to get it to find the target. Maybe you could have the person throw a
hat or coat at the professor before leaving. That way you have something to use as a
scent for the dog to follow. I also agree with Stephen's suggestion that the handler
should not know the correct response. That is probably very important.
I think you could turn this into a really fun experiment for a group of students to
work on.
Richard Platt
St. Mary's College of Maryland
