Every score has a
purpose. The problems with g.e.'s
have been well documented in this and previous discussions, but it is not that
there is something intrinsically evil with g.e.'s, but that they are often
misused and easily misinterpreted. Putting
g.e.'s in the hands of teachers has been compared to putting an AK-47 in the
hands of a child. A fine instrument,
easy to use (just pull the trigger), but generally you would want to have some
control over where and when it was actually applied.
HOWEVER, that said,
g.e's have an attribute that standard scores and percentiles do not.
If you use standard scores to document progress, you may easily discern a
drop from one testing to the next. In
an adversarial situation, the parents will LOVE those scores.
Indeed, if you review the background material for Brody v. Dare Cty, an
interesting due process lawsuit documented on Wrightslaw, the parents might even
convincingly allege (to a due process hearing officer) that a decline in
standard scores represents regression. The
results, if those scores are accepted as represented, can be catastrophic.
For example, in her
"Letter to a Stranger" the
parent argued that the following data proved regression:
Regressed: 1.5 years
of progress after 3 years of special ed.
(2.9 g.e. to 4.4.) Percentile Rank declined 12%
The Hearing Officer
ruled for the parents, providing them with tuition reimbursement.
On the other hand, in
HISD v. Caius, 2000, the court ruled that the school properly used grade
equivalents rather than percentiles to document progress.
The school argued for the use of g.e.'s, and it won.
The 5th Circuit, in finding that the school system had
provided FAPE, wrote "HISD employed the widely utilized and accepted
Woodcock Johnson intelligence and achievement test to indicate Caius's academic
progress. Caius's test scores showed the following changes from 1993 to 1995:
(1) Math scores improved from the 1.7 grade level to 3.1; (2) written language
improved from the 1.5 grade level to 1.9; (3) passage comprehension went from
1.7 to 2.2; (4) calculation rose from 1.4 to 3.3; (5) applied problems improved
from 2.0 to 3.0; (6) dictation went from 1.6 to 1.8; (7) writing improved from
1.4 to 2.6; (8) word identification, basic reading skills, and letter
identification rose from 1.8 to 2.1; and (9) word attack rose from the level of
a seven-month kindergarten student to grade level 1.8."
It went on to say, "From
1995 to 1996, Caius showed the following improvements: (1) Broad reading
increased from 2.1 to 3.3; (2) word identification from 2.1 to 2.8;(3) passage
comprehension from 2.2 to 3.9; (4) math from 3.1 to 4.4; (5) calculation from
3.3 to 5.0; (6) applied problems from 3.0 to 3.6; (7) written language from 1.9
to 2.9; (8) dictation from 1.8 to 2.8; (9) writing samples from 2.6 to 3.3; (10)
basic reading cluster from 2.1 to 2.8; and (11) proofing from 2.3 to 2.6. Only
word attack remained the same, at the 1.8 grade level."
In its decision, the court said, "Caius claimed that a child's
percentile scores were the best measure of academic performance, while HISD
argued that passing marks and advancement from grade to grade were sufficient
indicia to satisfy the IDEA. And on this dispute the district court is correct
that a disabled child's development should be measured not by his relation to
the rest of the class, but rather with respect to the individual student, as
declining percentile scores do not necessarily represent a lack of educational
benefit, but only a child's inability to maintain the same level of academic
progress achieved by his non-disabled peers."
Of course, since in
Brody v. Dare, the sped teacher apparently admitted she did not know how or have
the skills to teach reading, the issue of percentiles v. grade equivalents
probably wouldn't have made any difference. But it could have.
standard scores are best for demonstrating relative weaknesses and strengths.
G.E.'s, on the other hand, are most effective when trying to document
progress ("educational benefit" that is more than merely trivial.)