There was an interesting comment on this thread - assessing post-test
measures following training - on the list last night. Sorry, but I've
deleted the message and I don't remember who the author was. The comments
were:
the subjects aren't blind to which condition they're in, which
introduces significant bias in the outcome measure;
there are probably other significant sources of experimenter bias;
and giving a measure of "self confidence" as a pre-test introduces even
more bias, because it alerts the subjects to "what's wanted" in the
experiment (isn't there a non-reactive measure that could be used here
instead?)
I thought these were all important comments. But they get us further
off into questions of experimental design, and the original question still
hasn't been answered. So come on, guys and gals, which test is it? I
want to know.
Mike
