[eDebate] 100 point scale

David Glass gacggc
Mon Nov 5 19:04:32 CST 2007

yah, that's all I was advocating: clear instruction (and fingers
crossed that folks listen and follow the instruction)

because for every person who translates the new scale to A; 90-100, B:
80-89, etc,
there may be someone else doing the multiplication, and still someone
else continuing to use the top 2.5 points as if they're the entire

also, I'm not sure why you're surprised at the response, since the 30
point scale had clear instructions, and yet there has been gradual
point-creep towards the top.  so there at least seems to be a
legitimate risk that while some people will respond to the new scale
by broadening their range according to instructions, others won't.

and as you know, being a statistician, all you need is enough outliers
to make it into some teams judging sets more than others, and you will
then indeed have randomness...  and if there are 3 legitimate possible
models for people to follow - that makes it even more likely that the
scale wont broaden out synchronously


On 11/5/07, Gary Larson <Gary.N.Larson at wheaton.edu> wrote:
> A couple of quick observations:  I'm a bit bemused by our hand-wringing as
> to whether judges will be able to create appropriate standards or community
> norms in assigning points on a new scale (e.g. David's worry that it will
> take 20 tournaments before assigned scores prove to be more than random).
> Considering that the vast majority of judges also have the "day job" of
> assigning grades to undergraduate students that  help determine GPA's and
> can potentially impact graduate or job placements, I'm not encouraged to
> hear that we doubt whether they can handle a 100 point scale (or some other
> equivalent scale) for evaluating debates reliably.  Since we typically
> identify the participants in the debate community as among the brightest and
> best, I would hope that clear instructions about how point scales work would
> reduce the chance that some judges might think that the translation should
> be "old score + 70" while others think it is "old score * 3.3"  I would like
> to believe that clear instructions can be successfully interpreted by all of
> the various judges that we have in the pool.  But, of course, then perhaps
> we should imagine that judges could successfully modify their use of the
> current scale to reduce point inflation and scale compression.
> Any scale will ultimately succeed or fail based on the common understandings
> of bright people who are interacting with student performances.  That said,
> there is a different metaphorical force associated with a 30 point scale, a
> 50 point scale, a 100 point scale with whole points, a 10 point scale with
> decimal points, the current scale with the addition of quarter points, etc.
> Given my druthers as a sometimes statistician, I'm most easily convinced by
> the rhetorical merits of the 100-point scale since it has so many grading
> analogues in higher education.  But I hope that we're smart enough and
> committed enough to embrace a scale however we define it.
> I should also underline one concern about quarter-points as the author of
> tab software.  Considering the number of keystrokes required to enter number
> like 27.75 (while at present we can tab most scores with only 1 or 2
> keystrokes), IF we're going to improve discrimination by just adding more
> significant digits instead of using all of the whole (and half) points
> already at our disposal, my druthers would be to go to decimal points rather
> than quarter points.  And if we go there, then perhaps the best scale is
> 0-10 with decimal points.  But even there we have the experience of sports
> like skating and gymnastics relatively quickly discovering that they need to
> go to hundredths to get discrimination.
> _______________________________________________
> eDebate mailing list
> eDebate at www.ndtceda.com
> http://www.ndtceda.com/mailman/listinfo/edebate

More information about the Mailman mailing list