Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: grading schemes



I 'shoot' for a B- average on my tests (80-83% on my grading scale). If
that happens with the raw scores, I leave them alone. If the grades are
lower, especially if they are spread out over a wide range, then I like to
use the 'square-root' curve to adjust them. You can apply the curve on any
test by taking the square-root of the student score times the square-root of
the number of points on the test. For 100-point tests then a 25 becomes a
50, a 36 goes to 60, 49 to 70, etc. This curve has the property of
compressing the scoring range and skewing it towards higher numbers.

I use this because my overall grading scheme is a 1000 total point system
with grades determined by the total points earned. 930 or more points earns
an A, 900-929 an A-, etc. A single test that ends up too difficult (for
whatever reason) can pretty much wipe-out the A range without some
adjustment. The square-root seems to do the trick.

Rick

**********************************************
Richard W. Tarara
Professor of Physics
Saint Mary's College
Notre Dame, IN 46556
rtarara@saintmarys.edu

FREE PHYSICS INSTRUCTIONAL SOFTWARE
www.saintmarys.edu/~rtarara/
PC and MAC software
NEW! SIMLAB2001--DYNAMIC CARTS now available.
CD-ROMs now available
******************************************************

----- Original Message -----

At 19:41 -0500 12/17/01, Joe Heafner wrote:

How many of you assign (test) grades based on a "raw" score vs. some
kind of "scaled" score? By "raw" score, I mean the ratio of points
earned to total possible number of points. For example, if there are
60 possible points on a test and a students accumulates only 15
points, the raw score would be 25%, which is clearly a failing
performance (isn't it?). How many of you take these raw scores and
scale them somehow?