Hi, Michael --

In the research Roediger cited, the tests were free recall *with no
feedback*, and thus didn't supply any information that could later be
used to improve performance.  His suggestion is that the very act of
trying to recall things makes currently-un-recallable information more
likely to be recalled at a later test.

Imagine getting a list to learn, and then being given a blank sheet of
paper upon which you list items you recall, then being given another
sheet of paper, for a total of three tests.  No feedback, no chance to
learn from the test.  Over time (he re-tested up to a week later),
recall improves.  An alternate group was given a chance to re-study the
list rather than the testing.  Immediately, the many-study condition
showed better performance, but over time the many-test group surpassed
that performance.

So it isn't clear that this is teaching to the test at all.

And as for the UNM study, the practice tests did not (necessarily)
contain the items that were on the real test -- they were all sampled
randomly from a very large bank of questions.  I'd bet that some of the
items that appeared on the practice tests were on the real test, but I'd
also bet that it wasn't an appreciable number.  So it isn't clear to me
that the testing in that case contributed to specific knowledge of test
items, but it rather seems to have reinforced learning generally.

I do not think that anyone would dispute that testing yourself as you
study improves learning -- there are *lots* of data that show this.  I'm
just wondering how to structure this into a class, and whether it's a
good idea to try it.  

If I give the impression that I'm looking for a strategy to learn how to
better teach to the test, I apologize for not being clear.  I'm looking
for a way to increase retention in my classes.

m

-----Original Message-----
From: Michael Scoles [mailto:[EMAIL PROTECTED] 
Sent: Sunday, May 08, 2005 12:03 PM
To: Teaching in the Psychological Sciences
Subject: Re: Testing improves learning

It isn't too surprising that repeated "practice" tests improve
performance on the "real" test.  An obvious criticism of repeated
practice tests would be that they may contribute little to general
knowledge about the content area and more about specific knowledge of
the test.  A thoughtful discussion of this problem can be found at:

http://www.carnegiefoundation.org/perspectives/perspectives2004.Apr.htm



Michael T. Scoles, Ph.D.
Interim Chair, Dept. Psychology & Counseling University of Central
Arkansas Conway, AR 72035
>>> [EMAIL PROTECTED] 05/08/05 11:44 AM >>>

I'm wondering if any of you out there have read this literature and/or
employed frequent and repeated testing in a class, and if so, how.  He
spoke of someone (I've forgotten his name) at UNM who required a class
to take three on-line practice exams before the real one and compared
the results of that class to one for whom the three practice exams were
offered rather than required.  (And we can be pretty sure they didn't
take too much advantage of that.  Call my cynical again.)  The upshot of
that informal (quasi) experiment was that the class for whom the
practice was required scored about a letter grade higher than the other.


---
You are currently subscribed to tips as: [EMAIL PROTECTED] To
unsubscribe send a blank email to [EMAIL PROTECTED]

---
You are currently subscribed to tips as: [email protected]
To unsubscribe send a blank email to [EMAIL PROTECTED]

Reply via email to