[RUME] "Blanking" on tests: any research on this phenomenon?

[Henry Walker]

Here is another anecdote from 20+ years ago.  I think about this 
periodically, but have never determined just what to do with it.

In a regular Calculus I course, I gave three one-hour, in-class tests.
One student was consistently doing well on homework and in discussions
in the office.  However, for the first two tests, he consistently
performed terribly.  After each test, he indicated that he 
thought he understood the material rather well, but his written
tests were terrible.  Immediately after the third test, he came
directly to my office, indicating he had been similarly inarticulate
on that exercise.  He indicated he wanted to answer the questions
on my blackboard -- not for a better grade, but to prove to me that
he knew the material.  I had no objection to his standing at the
board while I did other work.  He proceeded to write an
essentially-perfect solution to each problem.  (When he finished
one, he interrupted me, so I could look at it.  After taking notes,
he went on to the next problem, and I went back to work.)

Since we had walked directly from the test room to my office,
it was clear that he could not have looked at a book or talked to
others.  Further, in considering the quality of answers, his
written work scored in the 50% range, while his board work was
over 95%.

With that experience, I allowed him to take the exam on my 
blackboard, resulting in a clear "A".  He took calculus II from
me the next semester,  Again, I allowed him to write his
answers on the blackboard rather than on paper; he put up a
solution while I worked on something else.  At his signal, I
looked at his work, and took some notes.  He then went to the
next problem, and I continued my other work.  

Overall, this suggested that there was something about taking a
math test in a written form in the classroom that caused him 
to "blank" or panic.  Taking the same thing on a blackboard
gave remarkably better results.

As noted earlier, I am not sure just what to do with this data
point, but the observation seems consistent and dramatic.

Henry Walker

-- 

Henry M. Walker
Samuel R. and Marie-Louise Rosenthal Professor
    of Natural Science and Mathematics
Professor of Mathematics and Computer Science
Department of Mathematics and Computer Science
Noyce Science Center
1116 8th Avenue
Grinnell College
Grinnell, IA  50112-1690
641-269-4208
Fax:  641-269-4984 and 641-269-4285
http://www.cs.grinnell.edu/~walker/