In general, I agree with Lise. It was my practice to have my answers available to students as they handed in their test papers, and as soon as the papers were graded (whether by me or by a grader) I would report a stem-&-leaf diagram on the blackboard (or overhead). I did not use any identification system at all, coded or otherwise: only the raw scores. Since my tests were invariably short-answer tests, and each item was marked out of 10 points (usually), the raw scores were not immediately interpretable in terms of letter grades. A class or two later, I would hand out printed copies of the stem-&-leaf diagram, with cutting points marked for conversion to letter grades.
As others have remarked, I think one owes a student enough information to be able to tell about how well s/he did with respect to the rest of the class. A S-&-L diagram does this, without identifying anyone else in the class: one has one's own raw score, on the returned paper, but one only knows another student's raw score if the other student chooses to communicate it. As for medians and means: well, that's part of what they're supposed to have been learning, so they can readily find the median and hinges for themselves, and with a little more work if they're interested they can find mean (and s.d.). I didn't usually use those statistics myself, so didn't usually bother to calculate them. And I consider it salutary for the student who knows that the score is low (but, perhaps for that reason, may not realize HOW low) to see the score near the top of a S-&-L diagram, and separated from the next larger score by a gap of three empty stems... I do not understand the issue of "hurt feelings" mentioned by some respondents. The hurt would be the greater, one would suppose, if there were no feedback for any test during the term, and only after the final exam a student discovered that s/he was failing, and had been failing all along. "Embarrassment" I can see, especially for one who is in the habit of getting As and finds oneself at B-minus unexpectedly, but "hurt"? On Sat, 18 Jan 2003, Lise DeShea wrote: > I agree with Dennis. In fact, I show a stem-and-leaf display on an > overhead projector -- when I have a class of more than about 15 > students. With a smaller class, they could figure out which student > had which score. I think it shows the value of graphing data and > reinforces an understanding of one kind of graph they have to learn > in my class. Plus, after collecting each quiz I immediately hand > out a copy of the key, so that they can learn from their mistakes > while the quiz is fresh in mind. By the time they see the > stem-and-leaf, it's not much of a surprise how they did. ----------------------------------------------------------------------- Donald F. Burrill [EMAIL PROTECTED] 56 Sebbins Pond Drive, Bedford, NH 03110 (603) 626-0816 . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
