IQ score differences across time may relfect real changes in the brain

Lay people and many professionals often express consternation when an individuals measured IQ scores are different at different times in their life.  This concern is particularly heightened in high stakes settings where differences in IQ scores can result in changes in eligibility for programs (e.g., social security disability income) or life-or-death decisions (e.g., Atkins MR/ID death penalty cases).

Factors contributing to significant IQ score differences are many (McGrew, in press a) and may include: (a) procedural or test administration errors (e.g., scoring errors; improper nonstandardized test administration; malingering; age vs. grade norms; practice effects), (b) test norm or standardization differences (e.g., norm obsolescence or the Flynn Effect; McGrew, in press b), (c) content differences across different test batteries or between different editions of the same battery, or (d) variations in a person’s performance on different occasions.

 An article "in press" (Neuroimage) by Burgaleta et al. (click here to view copy with annotated comments)  provides the important reminder that differences in IQ scores for an individual (across time) may be due to real changes in general intelligence related to real changes in brain development.  These researchers found that changes in cortical brain thickness were related to changes in IQ scores.  They concluded that "the dynamic nature of intelligence-brain relations...support the idea that changes in IQ across development can reflect meaningful general cognitive ability changes and have a neuroanatomical substrate" (viz., changes in cortical thickness in key brain regions).  The hypothesis was offered that changes in the the cortical areas of  frontoparietal brain network (see P-FIT model of intelligence) may be related to changes in working memory, which in turn has been strongly associated with general reasoning (fluid intelligence; Gf).

The cortical thickness-IQ change relation was deemed consistent with "cellular events that are sensitive to postnatal development and experience."  Possible causal factors suggested included insufficient education or social stimulation during sensitive developmental periods, as well as lifestyle, diet and nutrition, and genetic factors.

• McGrew, K. S. (in press a).  Intellectual functioning:  Conceptual issues.  In E. Polloway (Ed.), Determining intellectual disability in the courts:  Focus on capital cases.  AAIDD, Washington, DC.

•  McGrew, K. S. (in press b).  Norm obsolescence:  The Flynn Effect.  In E. Polloway (Ed.), Determining intellectual disability in the courts:  Focus on capital cases.  AAIDD, Washington, DC.

[Click on images to enlarge]

IQ Test DNA Fingerprints: Comparison of WJ III/BAT III to WJ-R/BAT-R

Here is another of IQ's Corner "IQ Test CHC DNA Fingerprint" test comparison series.  This particular CHC fingerprint figure compares the CHC composition of the respective full scale total composite IQ scores from the WJ III/BAT III and the the earlier version of these batteries....the WJ-R/BAT-R.

Background information regarding the development, use and interpretation of this IQ global IQ score feature can be found at a prior post and in the IQ Test CHC DNA Fingerprint section on the blog side bar.

I now present a comparison of the R/III versions of the WJ/BAT batteries as I have seen psych reports where a subject had previously been administered the WJ-R and was later tested with the revised WJ III (in the case of Spanish-speaking individuals, I've seen the BAT-R and the BAT-III---click here for more background information on the Spanish version of the WJ III...the BAT III).  

In the case of the WJ-R/BAT-R, the full scale IQ composite is called the Broad Cognitive Ability (BCA) cluster.  The name was changed in the WJ III/BAT III to General Intellectual Ability (GIA) cluster.  The name change was not cosmetic.  The use of the term "general intellectual ability" in the newest WJ III/BAT III reflects the fact that this global IQ composite score is designed to be the best statistical estimate of the theoretical construct of general intelligence (g) via the use of differential test weights.

Using principal components analysis, a g-factor was extracted from the seven WJ III/BAT III Standard Cognitive battery tests (at each age level), g-factor weights calculated (by age---they shift slightly as a function of age), and the g-weights used to differentially weight the contribution of the seven tests to the composite GIA-Standard cluster score.  The same process was completed for the 14 test GIA-Extended cluster score.  This procedure is explained in detail in the WJ III/BAT III technical manuals/reports and is also briefly summarized in a free on-line Assessment Service Bulletin technical abstract.

In the case of the WJ-R/BAT-R, the respective 7-test BCA-Standard and 14-test BCA-Extended cluster scores are based on the simple arithmetic average of each set of scores, thus resulting in an equally weighted global IQ score.

Thus, differences between the global WJ-R/BAT-R and WJ III/BAT III IQ scores may occur as a function of the respective scores reflecting differential contributions of the broad Gf-Gc abilities as per the CHC theoretical model that underlies the batteries.

Below is the IQ Test CHC DNA Fingerprint comparison of the two respective editions of the WJ-R/BAT-R and WJ III/BAT III.  The weights presented for the WJ III/BAT III are the median (average) weights across all age groups.  The previously referenced ASB (see above) includes a table of the specific weights by age.

[double click on figure to enlarge]

Although the CHC composition of the respective global IQ scores did not change dramatically, there are enough differences by CHC ability to suggest that slightly different global IQ scores may be produced for the same individual depending on whether they took the WJ-R/BAT-R or the WJ III/BAT III (assuming proper administration, scoring, etc.).  Consistent with psychometric intelligence theory (aka., CHC theory), the WJ III/BAT III global IQ scores (GIA-Stnd; GIA-Ext) are more heavily weighted as per a subjects performance on the more g-loaded measures of Gf (fluid intelligence/reasoning), Gc (crystallized intelligence or comprehension-knowledge), and Glr (long-term storage and retrieval).  In contrast, abilities that are less cognitively demanding and more related to perceptual (Gv, Ga), speed (Gs), and short-term memory (Gsm) functioning contribute slightly less to an individuals WJ III/BAT III global IQ GIA score than was the case with the WJ-R/BAT-R.

If significant differences are found when comparing scores from the respective R/III editions of the WJ for an individual, examiners should review the Gf-Gc CHC test/cluster profiles to determine if some (or all) of the score differences might be related to the shift from an equally weighted global IQ score (WJ-R/BAT-R) to a differentially-weighted (WJ III/BAT III) global IQ score.  In theory, an individual could obtain very similar test-level scores on each battery, but because "all scores are not created equal" (in the estimation of general intelligence or g) in the case of the WJ III, a shift in the global GIA IQ scores may occur.

Other IQ Test CHC DNA Fingerprint comparisons can be found by clicking here.  More will be coming in the future.

[Conflict of interest note:  I am a co-author of the WJ III/BAT III]

Technorati Tags: Psychology, school psychology, educational psychology, forensic psychology, neuropsychology, special education, intelligence, cognitive abilities intelligence theories, CHC theory, Cattell-Horn-Carroll, factor analysis, cross battery assessment, IQs Corner, IQ, IQ tests, IQ scores, Gv, Gc, Gf, Ga, Glr, Gsm, Gs, Gq, WJ III, WJ-R, WJ III NU, BAT III, BAT-R, Woodcock-Johnson, Bateria, Spanish-speaking, Hispanic, bilingual IQ testing, IQ Test CHC DNA Fingerprints

IQ Test DNA Fingerprints: Comparison of WJ III/BAT III, WISC-IV and WAIS-IV

Here is another in the series of IQ's Corner "IQ Test CHC DNA Fingerprint" test comparison series.  This particular CHC fingerprint figure compares the CHC composition of the respective full scale total composite IQ scores from the WJ III/BAT III and the WISC-IV and WAIS-IV.

Background information regarding the development, use and interpretation of this feature can be found at a prior post and in the IQ Test CHC DNA Fingerprint section on the blog side bar.

More to come. [Double click on image to enlarge]

Technorati Tags: psychology, school psychology, educational psychology, clinical psychology, neuropsychology, forensic psychology, forensic psychiatry, intelligence, IQ tests, IQ scores, IQ, IQ testing, WISC-IV, WAIS-IV, WJ III, WJ III NU, BAT III, BAT III NU, applied psychometrics, psychometrics, IQ Test CHC DNA Fingerprints, IQs Corner, Atkins cases, death penalty, or punishment, mental retardation, intellectual disability, MR, ID, special education, learning disabilities, LD

Why do IQ scores often differ? Check out IQ's Corner "IQ Test CHC DNA Fingerprint" analyses

Understanding IQ score differences possibly related to differential CHC content difference coverage.  See ICDP blog post where it is explained and illustrated.  where new feature is introduced - IQ Test DNA CHC Fingerprints

[Double click on image to enlarge]