Hey, Albert --

If you have classes that small and had students who scored well, you
surely have enough variability to suggest that there aren't any real
differences in performance between those classes.  I'd treat them all
the same.

In large classes, a mean of about 2/3 was pretty common for me -- I'm
not sure what that says about my teaching and/or my exams, but that's
what I shot for.  There are always other assignments that factor into
grades to bring the course grades up from there, so I wouldn't at all be
alarmed at those scores.  I wouldn't adjust them -- although I might
make some adjustments at the end of the semester.  In very large classes
(100+) I graded relatively (using z-scores, but with a "minimum
competency" requirement), and that tended to help the final grades
somewhat.

The suggestions you got about item analysis are the best.  If there are
bad items in there (but a low _d_ -- or whichever stat you can get for
item analysis -- doesn't always mean it's a bad item), you can throw
those out and adjust the grades based on the number of surviving items,
and then let the chips fall where they may.  This
<http://www.uleth.ca/edu/runte/tests/itemanalysis.html> looks like a
nifty little website for doing it if you've not done it before.

But I'd think twice about making an adjustment simply on the basis of
their performance on this exam. I don't know about the structure of your
course and what other opportunities the students have to show what
they've learned, but being too quick to make adjustments as you go along
can have bad (grade-inflationary) consequences down the road.

m  


------
"There is no power for change greater than a community discovering what
it cares about."
--
Margaret Wheatley 

-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] 
Sent: Monday, November 12, 2007 4:30 PM
To: Teaching in the Psychological Sciences (TIPS)
Subject: [tips] Question about Exam Scores


Dear Colleagues:
 
I teach at the community college, and this semester my teaching load is
four courses.  Today, I administered the second exam in all of my
courses and just scored each of them.  The average for each class is as
follows
 
Class 1- N=27 70% average raw score
Class 2- N=31 66% average raw score
Class 3- N=29 67% average raw score
Class 4- N=30 67% average raw score
 
The exam consisted of 50 multiple choice items and scored on a 100%
scale.  My colleagues at my campus are divided, some say that they
deserve the score they get, others will curve the exam scores.  In fact
several of my students asked if I would curve even before I handed out
the answer forms and exam booklets. 
 
Each semester I do have a handful of students who do not test very well,
however this semester I seem to have more of those that do not test well
or are not studying adequately for my exam.  In each class three
students scored 90% or higher  
 
My question is as follows
 
At what point does one scale or curve the results and are there any
specific methods of doing so?  
 
Thanks
 
Albert Bramante
Department of Psychology/Sociology
Union County College
[EMAIL PROTECTED]




________________________________

See what's new at AOL.com <http://www.aol.com?NCID=AOLCMP00300000001170>
and Make AOL Your Homepage
<http://www.aol.com/mksplash.adp?NCID=AOLCMP00300000001169> .

---
To make changes to your subscription contact:

Bill Southerly ([EMAIL PROTECTED])

---

Reply via email to