On Sat, 18 Jan 2003 21:25:07 -0500, Stan Brown
<[EMAIL PROTECTED]> wrote:

> Stan Brown <[EMAIL PROTECTED]> wrote in sci.stat.edu:
> 
> >After most quizzes, I show the class the median and mean. I have 
> >sometimes thought about displaying the further information mentioned 
> >above, but have desisted exactly because I'm concerned about hurt 
> >feelings.
> 
> I've been reading people's comments with interest. The consensus 
> seems to be that it will do no harm to display a list of grades, 
> properly coded so that students can't infer each other's grades.
> 
> On the other hand, are there actually any positive benefits to doing 
> this? Does it actually give students enough _useful_ information to 
> justify the extra effort?

Quality control?

In 1966, I complained to a psychology professor that the grading
on the second Hour exam, which was in sentences, seemed to have
been shoddy.  The exam entailed   describing of research studies in 
the Readings.  People I knew who didn't study beyond reading the 
abstracts  scored better than people, including me, who had read 
every article. 
 - My working hypothesis was that   TAs  who did the grading 
had only read the abstracts. 
 - My further evidence  was the negative correlation between 
the 1st Hour exam (multiple choice)  and the 2nd.  I could document
that because the scores for both exams were still posted.
 - The result of my complaint was that the  scores for the 1st exam
were immediately taken off the bulletin board.


By the way, last year, some mother took a school into court 
because her child was being shamed by having rotten papers 
handed back in public, in class.   The mother lost.

--
Rich Ulrich, [EMAIL PROTECTED]
http://www.pitt.edu/~wpilib/index.html
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to