Since 2002, 4 million visitors plus:
hit counters
search engine optimization service

  Appletcollection Vertical Menu java applet, Copyright 2003 GD

TEST SCORE LABELS - The Hieroglyphics of Special Education

As one who "made the top half of the class possible," I would add that "average" is determined arbitrarily by the publishers or authors of tests.  For example, a standard score of 110 is "average" on the WJ III and most Pro-Ed tests, but High Average on the Wechsler scales and Above Average on the DAS. 

The Manual for the VMI 4th ed. offers several classification schemes, but the one that appears to be official calls "average" scores from 83 through 117!

A standard score of 83 is "Average" on the VMI, "Low Average" on the Wechsler scales and WJ III, and "Below Average" on the DAS and most Pro-Ed tests.

Many speech and language tests define the "average range" as 85 - 115, rather than 90 through 109 or 110.

"Well, Mr. and Mrs. Parent, we were a little concerned about Sally's Low Average standard score of 85 on the WISC-III Verbal IQ, so we administered the Verbal scale of the DAS and, sure enough, Sally scored 85, which is Below Average.  Her 85 on the DAS was lower than her 85 on the WISC.  We thought it might be a language problem, but the speech pathologist reports that Sally's scores of 85 were all Average after all in spite of the discrepancy between the Low Average WISC 85 and Below Average DAS 85."

Most tests use the term "scaled score" for subtest standard scores with a mean of 10 and s.d. of 3, but Pro-Ed calls them "standard scores" and usually uses "quotient" for standard scores with a mean of 100 and s.d. of 15. 

"Also, Mr. and Mrs. Parent, Sally did much better on Story Construction with a standard score of 11 than she did on the WISC Verbal Comprehension with a standard score of 86."

Some evaluators carefully translate all scores into a single scoring system (such as one Cathy Fiorello offered on this list some time ago; or the one proposed in
Kevin McGrew, Dawn Flanagan, Sam Ortiz, et al.'s Desk References and other books on their integrated CHC Cross-Battery Approach; or stanines).  If we do make such translations, we need to keep reminding readers we have done so and need to append explanations of the original classification systems and our own system to the report.  If we stick with the original scoring system for each test, we need to explain the scoring systems very clearly.

What we must not do, IMHO, is believe our own classification levels.  If a score of 79 suggests a possible problem, the problem has not disappeared if the score soars all the way to 80. A standard score of 85 is still at the 16th percentile no matter what we call it.  I have heard evaluators report that they had really thought the child had a problem, but everything turned out to be OK because a score of 85 was "average" after all.

It is, I think, very good practice to use a 90% or 95% confidence band and report all the classification labels intersected by the band (e.g., "Below Average to Average" or "Borderline to
Superior").


John Willis