Wednesday, March 09, 2005

IQ scores, NCLB & Forrest Gump: Run Forrest..Run

As a professional who has written about intelligence theories and tests and who is a coauthor of a frequently used individual measure of intelligence (WJ III), I often find other professionals and educators shocked when I highlight the less than perfect predictive capability of IQ tests. Although test manuals and research reports often report high correlations between IQ and achievement tests (e.g., .60-.70's), I occassionaly hear statements that suggest that some educators and assessment professionals fail to recognize that such high correlations (although some of the highest in all of psychology) are evidence of the fallibility of intelligence tests--they are less than perfect predictors.

Correlations of this magnitude tell us that IQ tests, on their best days, predict 40-50% of school achievement (
Applied Psychometrics 101 – square the correlations and multiply by 100 to get the percent of variance explained). This is very good. Yet….50-60% of a person’s school achievement is still related to factors “beyond IQ!”

An unfortunate unintended negative side effect of the success of IQ tests can be the implicit or explict use of global IQ scores to form substandard or low expectations for individuals. How often have we all heard someone state, after hearing a child's general IQ score that, after using some norms or forumula to generate an "expected" achievement score, that teachers and parents should expect the child to achieve "at or below" these already below average expectations? It is often not recognized that for any given level of IQ score, half of the individuals with that score will achieve at or above predicted levels of expected achievement (based on that score). In the context of NCLB (No Child Left Behind), there is a real fear that IQ test scores may seduce educators and other education-related professionals into the
“soft bigotry of low expectations (it was either G. W. Bush or his then Education Secretary who coined this phrase).

The National Center on Educational Outcomes (NCEO) has recently published a report dealing specifically with this issue in the context of NCLB.
Expectations for Students with Cognitive Disabilities: Is the Cup Half Empty or Half Full? Can the Cup Flow Over? (McGrew & Evans) is a report that addresses this issue. This report is pubished on the NCEO web page (click here). [If you go to the following page(click here) and right click on PDF (after the report title) you can download a pdf copy to your hard drive.] --- isn't technology wonderful?

As stated in the report introduction,

This report…includes an analysis of nationally representative cognitive and achievement data to illustrate the dangers in making blanket assumptions about appropriate achievement expectations for individuals based on their cognitive ability or diagnostic label. In addition, a review of research on the achievement patterns of students with cognitive disabilities and literature on the effects of teacher expectations is included. The literature raises numerous issues that are directly relevant to today’s educational context for students with disabilities in which both the Individuals with Disabilities Education Act (IDEA) and the No Child Left Behind (NCLB) Act of 2001 are requiring improved performance. Particularly for those students with cognitive disabilities, the information on expectancy effects should cause us much concern. Is it possible that expectancy effects have been holding students back in the past? Are we under the influence of silently shifting standards, especially for students with cognitive disabilities? It is anticipated that the information in this report will help guide decisions about appropriately high and realistic academic expectations for students with cognitive disabilities.

The fictitious story of
Forrest Gumpis used in the report to illustrate the potential danger IQ score based generalized low expectations for students with disabilities---food for thought for educators, parents, and professionals involved in the education of students with disabilities during the current wave of NCLB-driven education reform.

After reading the report readers should feel compelled to yell
“run Forrest run….from the potential negative impact of the soft bigotry of low expectations.

3-14-05 note. Please note, in the spirit of full disclosure and potential conflicts of interest, that I was a coauthor of the NCEO report. I received $$ for writing the report but receive nothing for the number of copies that will be printed or distributed. Sorry for not pointing this out in the first post. I'm new at steps.


Anonymous said...

I must say, I was rather surprised when I visited the NCEO link to discover you were the co-author of the report. Returning to your log, I see that, yes, it does say McGrew & Evans in parenthesis. Didn't catch that the first time. But otherwise, it reads as if it was something you had read, sir, not written.

Anyway, this is fun! Thanks!

Kevin McGrew said...

Thanks for the observartion and feedback. I have now edited the original post (at the bottom) to reflect any potential conflict of interest. I'm just learning this new steps.

Anonymous said...

Okay, I've gotten to page 11 of the report. It's an interesting discussion of the WJ norming data. But here's a little reality twist--the kids who have IQ's between, say, 60 and 80, but who achieve in the average range, DON'T GET REFERRED TO SPECIAL ED. Kids who get referred have, I would estimate, two years of teachers trying everything they can to promote normal achievement. Two years of frustrated children, parents, and teachers. All frustrated by the attempt to achieve normal expecations. That is when the testing takes place, and when the adults can say, "Oh, I see. This kid's IQ is only 81. No wonder they only get C's and D's despite everyone's hard work." It's the facts on the ground that create the expecations for these kids, not the IQ testing.

Robert Misak said...

But then again, once the teachers/principal/parents see the "80", they figure "Oh, than that is what I should expect." That is one of the beauties of the application of CHC to evaluations; some of the time time the student is not a "Straight 80". I just consulted on a third grader with a 75 on the WISC-IV. When viewed under the CHC and additional tests were given (via cross-battery), this young lady turned into a student with 5/7 broad cogntive abilities within normal limits (and Auditory Processing >115) and a Short-Term Memory in the Very Low range and Crystallized Intelligence in the low 80's. If the educational diagnostician would have gone to the IEP team with the original canned report/interpretation, they would have looked at that 75 and said "Oh, no wonder she is failing! She is borderline retarded!" and she would have been placed in a self-contained setting. But now, they are looking at specific interventions focused on remediating/compensating her Gsm and Gc defecits. This is one of the problems with viewing FSIQ as a predictor of academic success. Just my 2 cents.

Anonymous said...

I have no disagreement with anything you say, Cognimetrixx. When an evaluation is done, it should be done thoroughly and not simply produce a single number. Agreed. I'm still skeptical that low IQ scores are producing low expectations. I still think the kids with low IQ scores who get evaluated are already having achievement problems that have nothing to do with low expectations.

Anonymous said...

I think that the WISC-IV no longer is a measure of g (i say that with many reservations already noted about g; please be kind), but rather a measure of a number of disparate cognitive functions. Thus the Full Scale IQ would be more appropriately labed "Test Composite," as it is for one of its competitors. I have found many learning disabled children to have high subtest scores for similarities, block design, matrices, and vocabulary, but very low full scale iq scores.

I like Reid Lyon's and Joe Torgesen's findings that, for young children, rapid naming and phonemic awareness are better academic predictors. Teachers often want an IQ score for children exiting preschool sped and I struggle to reframe these discussions.


Anonymous said...

in the interest of self-discolure of potential conflicts of interest, i have no lucrative consulting clients and take no pharmaceutical money, but will gladly consider any and all offers.

Kevin McGrew said...

Steve . LOL (laugh out loud). I love your satirical self-disclosure statement.

Unknown said...

Hey Kevin,I really liked your blog. It reflects quite a lot about IQ. IQ Tests are a good way of analyzing a person's intelligence, but it is definitely not the best ultimate one. In this regard I visited many sites offering IQ tests. Of all I was quite impressed with IQ Test

Research Papers said...

Many institutions limit access to their online information. Making this information available will be an asset to all.