At 01:30 PM 1/12/01 +0000, Gene Gallagher wrote:
No, Each one of the 1539 schools in the state was evaluated. The
districts were simply sent the results on the percent of their schools
that had failed to meet their targets or that had met their target
increases. All schools had to increase their scores, with the poor
schools being expected to improve by 6 points and the best schools by 2
points.
=========
surely, some schools are larger than others .... so, some school means are
based on many more classes than others but, in any case ... the mean of a
school will be based on 50 kids ... 100 kids ... or more? and i bet that
these schools ... even when based on 50 kids ... or 100 kids ... their
MEANS will not change much (and even if their mean improves ... that does
not mean their relative postion in the overally 1539 will change ... any )
... unless of course, the next year's 4th graders somehow ... are radically
different .. that is, while the current 4th grade looked like X ... the
INcoming 4th grade (last years 3rd graders) is either much more able ... or
less able
in any case ... unless we have the correlation between the two columns of
1539 means ... for schools ... for the two years ... again, we cannot speak
about regression to the mean ...
and from what you have indicated ... we have scaled scores ... so is it the
target to change in scaled scores??? or actual raw scores on tests? if it
is based on scaled scores ... we have a big problem ...
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================