RE: grading

I'm not sure that this is a criticism of MC questions so much as it is a
criticism of poorly constructed MC questions.

It is relatively easy to build MC questions that can only be solved by the
student working through the same processes that s/he would were s/he to be
answering an essay question. The trick is to focus on application rather
than definition. Most test-bank MC questions do the latter.

Antony Davies

-----Original Message-----
From: Michael Tamada [mailto:tamada_at_oxy.edu]
Sent: Tuesday, December 05, 2006 7:51 PM
To: Helen Roberts; Martha L. Olney; Bill Goffe
Cc: Teach-Econ List
Subject: RE: grading

Similar to Prof. Olney, I never use multiple choice exams, but I have the
luxury of usually teaching small classes. By requiring that the students
show their work, I can see their thought processes and point out where they
went wrong, if they make a mistake. And also give out appropriate partial
credit. However I have no actual evidence that this sort of grading (and
teaching) is actually more effective than the seemingly superficial (but
perhaps equally effective?) use of multiple choice exams. Certainly with a
large classroom, the close grading of indvidual exams becomes impractical,
and multiple choice exams may be the only way to survive.

I don't literally use the MC plus comment technique, but I do follow the
Univ of Chicago's technique of asking several questions which are "true,
false, or uncertain -- and explain why". But I make sure the students know
that the T/F/U answer is almost irrelevant, what matters is the "explain
why" part.

I guess one reason why I am skeptical of multiple choice exams is because as
a student I was extremely good at them -- even when I darn well knew that I
didn't really know the material. For awhile that was merely my personal
view, but about 20 or 25 years ago there was an excellent article in the
Atlantic which criticized the SAT. The article included I think 5 actual
questions from the verbal section of the SAT, the kind of question where you
had to read a passage and then answer questions about the passage (e.g. "The
author of this passage would be most likely to disagree with which of the
following statements?"). Well the article showed the questions and the
multiple choices of answers -- but they deliberately left out the passage to
read! I.e. all one could do was look at the question, the five possible
answers, and guess. I got 4 out of 5 correct, and the author of the
Atlantic article noted that most of his fellow reporters (alums of good
colleges mostly) also did quite well on this little exam.

With that, one has to wonder to what extent a multiple choice exam measures
actual knowledge, compared to ability to game a test or master test-taking
skills.

There's no question that multiple choice test such as the SAT are indeed
correlated with some measures of academic ability and/or intelligence. But
it's the strength of the correlation that I wonder about.

--MKT

-----Original Message-----
From: Helen Roberts [mailto:hroberts_at_uic.edu]
Sent: Tuesday, December 05, 2006 4:06 PM
To: Martha L. Olney; Bill Goffe
Cc: Teach-Econ List
Subject: Re: grading

I've found MC plus comment to be a useful way to get coverage plus
thinking. Students get one point for the correct MC answer and up to 2
additional points for their comments.

Using rubrics, it doesn't take me much longer to grade written exams than
MC exams.

I also make all my principles students decode/analyze/apply/evaluate
something relevant from the media like a newspaper article on every exam.

Helen

At 04:36 PM 12/5/2006, Martha L. Olney wrote:

>I'm spoiled (see, I acknowledged it). . .I have 1 TA for each 50
students
>in principles and 1 TA for each 60 students in upper-division courses.
So
>my exams are all written exams, no MC.
>
>One of my colleagues has gone to a mix of written & MC, as a way of
>covering more material. I haven't, as the way in which I teach doesn't
>seem to me to lend itself to MC questioning.
>
>Also, I feel strongly that I want the students to leave the class with
>some "take-away messages." Those messages aren't going to be from a MC
>question about elasticity, but rather (I think) from the experience of
>applying a concept to a real-world situation they are likely to
encounter.
>
>I was pleased this term in my upper-division Econ Hist class, when I
>mentioned a midterm question from my Spring 05 Principles class, that a
>number of students smiled in recognition. Both times I was talking
about
>how the closing of the gap between men and women in comparative
advantages
>between market & home production had, over the last 40 years, led to
less
>marriage and later marriage (less specialization and trade). Same
point,
>same concept, different classes. I'm going to bet that Sibo Zhao (one
of
>the students who was in both classes) remembers that point for a long,
>long time.
>
>That said, I drag out grading far longer than I should... I've got 80
>3-page papers waiting to be graded and instead I'm writing to you all.
>
>:-)
>
>Martha
>
>=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=
>
> Martha L. Olney
> Adjunct Professor of Economics
>
>University of California MOlney_at_econ.berkeley.edu
>Department of Economics http://socs.berkeley.edu/~olney
>549 Evans Hall, #3880 Office: 510-642-6083
>Berkeley CA 94720-3880 Fax: 510-642-6615

**********************************************************
Helen Roberts
Clinical Assistant Professor and Associate Director
UIC Center for Economic Education
601 S. Morgan #2103 (m/c 144)
Chicago, IL 60607-7121
Phone: 312-355-0378
Fax: 312-996-3344
E-mail: hroberts_at_uic.edu
**********************************************************
Received on Tue Dec 05 2006 - 20:00:09 EST