SK> But the longer this thread goes on, the more I think about it, and the
 SK> more I think that "the right mix" is the wrong idea.
 SK> I think it simply boils down to this: A few trouble makers in ANY size
 SK> class, including a small class, can really throw a wrench into the
 SK> works. When this handful of trouble makers is removed, (say they happend
 SK> to be absent on the same day) the class becomes positively enjoyable.
 SK> What this boils down to, then, is an argument in favor of
 SK> self-motivated, cooperative students.
Exactly.  I have taught classes ranging in size from 1 to 39.  In
general, I find that the "best" number ranges from about 18-22.
Larger classes, 23-29 can work if there aren't too many problem
children or they can be effectively stifled.  OTOH, when you get beyond
22, the middle to lower level kids (on the motivational scale, if not
the academic scale) tend toward feeling anonymous.  There are enough
other personalities, good & bad, in the room to deflect the teacher's
attention from them, so they feel empowered to daydream or mess around.
In fact, I've had a few who participated in shenanigans transferred to
other periods with smaller enrollment.  It appeared that the combination
of splitting them from their partners-in-crime & being able to increase
the amount of supervision I can give them in a smaller class contributes
somewhat to their progress.  They didn't become model students by any
stretch, but they managed to pull a passing grade.
Removing the "troublemakers" does seem to be a key, but in really large
classes (over 30 students,) the ability to disappear in a crowd can
enable those borderline students, who might otherwise stay on task, to
be somewhere else mentally & lose the value of instructional time in
much the same way as when a troublemaker disrupts teaching.
 SK> Given, then, a mix of students of the same "cooperation levels", I would
 SK> bet that class size definitely DOES make a difference. As a matter of
 SK> fact, this is one of the things that bothers me so much about studies
 SK> which purport to find evidence in one direction of the other about how
 SK> class size affects performance. The fact is that there are SO MANY
 SK> variables in any group of students that affect their performance, I
I've always wondered where these studies were done, you know, the ones
which said 35 students in a kindergarten class is okay...?  I've taught
kindergarten, & NO ONE should be confined with more than 15 of the
little devils.  They move so darned fast.
 SK> don't see how we can control effectively during any study to be certain
 SK> that any differences in performance can be reliably attributed to that
 SK> single factor: class size. District wide studies, which involve large
 SK> numbers of students such that minor differences such as this would tend
 SK> to be negated would have more merit IMHO.
Certainly, one would hope that any entity producing such a study would
have respectable credentials.  IMO, self-appointed experts who have NO
classroom experience should not be making pronouncements about how 60
college students are more difficult to manage than 60 elementary aged
children; and most especially obnoxious are the quasi-experts who tell
one to "ask any _______" (fill in blank with job or profession only
emphemerally relevant to the expert's argument.)  But that's only MY
opinion.
Leona Payne
... Ask any mermaid you happen to see, "Who's the biggest tuna?"  MS of NC.
--- Via Silver Xpress V4.3P SW12194
---------------