The T-Maze and Effect of Music on Learning Ability of Mice (again.)

Bill Skaggs skaggs at bns.pitt.edu
Sun Nov 14 19:11:10 EST 1999


edchong at my-deja.com writes:

> Hello everyone,
> I'm a 15 year old student doing a project on the effect of music on the
> learning ability of mice. I am currently facing two problems. I know
> this message is long but if you know anything about the T-maze or
> animal/mice behaviour, please help. I desperately need it, because I am
> running out of time and my experiment seems to have reached a major
> roadblock.
> 
>  [etc . . .]

First, you shouldn't feel too bad because you're running into
difficulties.  Whenever you're working with behaving animals, it's
very difficult to know whether a training procedure is going to work
until you've actually tried it.  I've been training rats to do things
for years now and they still constantly surprise me.  In fact,
granting agencies usually won't provide funding to do an experiment
unless there is preliminary data, because there's just no other way to
be sure that the experiment will really work.  Part of the reason for
doing the sort of experiment you're attempting is to learn by hard
experience the difference between a thought-experiment and an
experiment that can actually be carried out.  It's always a painful
lesson to learn -- it certainly was for me.  

Second, I think the design of your experiment is statistically valid.
In my experience it's difficult to get rodents to pay attention to
sound cues (except the sound of ripping paper, which freaks them out,
maybe because it sounds like a cat's hiss), so I share your teacher's
doubt that you'll actually find anything, but if you did somehow
come out with a statistically significant result, I think you would be
able to believe it.

Or at least, you'll be able to believe that the music made a
difference to the mice's performance.  Whether you can attribute any
observed differences to learning might be another question.  The
T-maze is actually a much trickier apparatus to use than it first
seems, and there are many factors that can affect the way an animal
behaves in it.  In particular, mice that are calm and "relaxed" have
a strong tendency to explore the side of the T-maze opposite to the
one they most recently visited, but mice that are frightened tend to
go to the same one they were last at (which maybe seems safer to
them).  Now, mice are very skittish animals, so any differences in
handling the groups could potentially have large effects on their
behavior. 

Normally, if I were going to try an experiment like yours, the first
thing I would do is spend several hours handling each of the animals,
holding it and letting it crawl around on your hands and arms, etc,
until it is used to you enough that it doesn't seem frightened when
you pick it up.  (I'm told that this is harder to do with mice than
rats).  This habituation process is excruciatingly boring, but it
helps tremendously.  It's much better to handle them with your bare
hands than to use gloves, though there is some risk of getting bitten
at first, and you should make sure you've had a recent tetanus shot if
you're going to do this.  The first few times you pick up an animal,
pick it up by the tail, but once it is familiar with you, you should
be able to pick it up by the body or just let it climb onto your
hand.  Feeding the animal while you're handling it also is helpful in
letting it get comfortable with you.

The next step is to train the animals to do some
other food related task, related to the one your experiment requires
but not so similar as to be confusable with it (for example, exploring
for scattered food in a small open box).  This gets the animals used
to the idea of working for food, and animals that seem particularly
stupid can be discarded at this stage without violating the
experimental protocol.  When the animals do well at the preliminary
task, you're ready for the actual experiment, and you should try to
keep the environment as non-frightening as possible.  It is good, as
somebody else wrote, to use a small apparatus, and keep the area
around it as limited as possible, and keep the light level as low as
you can and still see what's going on.  (Rodents prefer darkness.)

Now, being in the middle of your experiment already, you probably
can't do most of these things.  If you can give the mice some extra
handling, though, you probably should -- just be sure to handle all
groups equally.  You can also probably do some things to reduce the
frightfulness of the environment.  It may not be possible to rescue
the experiment at this stage but you can take a shot at it.

I wish you the best of luck,

	-- bill Skaggs





More information about the Neur-sci mailing list