Perceptual speed (Gs-P): Time to revisit CHC test classifications?

I recently read an article by Ackerman and Beier (2007) that reminded me of something I wrote in my Cattell-Horn-Caroll CHC (Gf-Gc) Theory: Past, Present and Future CIA book chapter (click here for on-line earlier version of the chapter; click here for specific section of the chapter I'm referencing).

The literature I reviewed in 2004/2005 suggested that those of us involved in the development and interpretation of intelligence tests should recognize an updated taxonomy of human speed abilities. A tentative taxonomy was presented, based on the work of others (I was just the synthesizer), that included Broad Cognitive Speed (Gs), Broad Decision Speed (Gt), and Broad Psychomotor Speed (Gps). Of particular interest to me was the suggestion that Perceptual Speed (P) may better be conceptualized as an intermediate speed ability between the broad (stratum II) and narrow (stratum I) speed abilities.

The research of Ackerman et al. suggested that the domain of Perceptual Speed might be best broken down into the sub-abilities of Perceptual Speed:Pattern Recognition (Rpr), Perceptual Speed: Scanning (Ps), Perceptual Speed: Memory (Pm), and Perceptual Speed: Complex (Pc). Definitions of these four narrow P abilities can be found at the 2004 link above, as well by visiting one of the original Ackerman et al. (2000) articles where this structure was first articulated.

In 2005, in an updated/revised CHC classification of the WJ III tests, I classified the WJ III Visual Matching and Cross Out tests as possible Ps measures. I further suggested that Pair Cancellation may be primarily a Pc test.

I would like to lay down a challenge to my fellow CHC intelligence users - how about people attempt to reclassify all P tests across the major intelligence batteries (as listed in the extensive CHC Cross-Battery publications of Flanagan et al.) as per the Ackerman four sub-P abilities? Maybe the results might help us better understand why performance on P tests often vary within test profiles and research studies. If nothing else, I'd like some feedback, challenges, revisions, suggestions, etc. to my reclassification of the three WJ III tests listed above.

Just a "Friday-afternoon-stuck-in-my-home-office-due-to-a-late-winter-storm-in-MN" idea.


Technorati Tags: psychology, school psychology, educational psychology, neuropsychology, Perceptual Speed, Gs, cognitive speed, CHC, Cattell-Horn-Carroll, intelligence, IQ, IQ tests, cross-battery, WJ III, Woodcock-Johnson

Processing speed (Gs) developmental growth on two WJ tests

I just skimmed an interesting (an largely theoretical) article by Kail and Ferrer (2007; Child Development-click here to view) that fit different mathematical models of age-related (developmental) growth to the WJ-R (Woodcock-Johnson--Revised) Visual Matching and Cross Out tests. The article does not have immediate practical applications. I'm was interested in the article since it deals with two tests from the WJ III (conflict of interest note - I'm a coauthor of the WJ III). I also read the article because of the first author...Kail. He, IMHO, is one of the top researchers investigating the nature of developmental growth of cognitive processing speed (Gs).

The abstract for the article is below.

The authors concluded that quadratic and exponential models fit the growth patterns of the Visual Matching and Cross Out tests the bests. The quadratic model was the best fitting model. The authors use this finding to illustrate how such analysis might be useful in exploring the mechanisms that underlie growth in cognitive speed (Gs).

For example, the authors note that the "parameters of these quadratics are often qualitatively like those obtained here: nonlinear change is achieved from a linear increase coupled with a nonlinear (power function) decrease." They then point other physiological/cerebral functions that show similar quadratic growth patterns - e.g., total cerebral volume and total body fat all show the same pattern of quadratic change in childhood and adolescence. The authors suggest that such a finding may suggest that all might have a "common (unspecified) biological base."

Interesting stuff...but not all that practical for the field of applied intelligence testing. However, I think the finding that the two tests showed the same pattern of developmental growth might be interpreted to support the interpretation of the two Gs tests as measuring the same underlying cognitive construct.

As I've written elsewhere, John Horn frequently talked about different types of validity evidence of human ability constructs--structural (factor analytic), genetic or heritability, neurocognitive, criterion-outcome, and developmental. We know from EFA/CFA (structural) studies of the WJ-R and WJ III that Visual Matching and Cross Out load on a common broad Gs factor. Logical content analysis has suggested that they are both measures of the narrow Gs ability of P (perceptual speed). The finding that both VM and CO display the same longitudinal developmental growth pattern, when combined with the extant EFA/CFA structural research that finds these two tests always loading on a common factor, in my opinion supports the validity of the logical narrow (stratum I) classifications of both tests as measures of Gs-P.

Just my two cent applied/practical extraction of information from this largely theoretical piece of research.

Abstract

• The primary aim of the present study was to examine longitudinal models to determine the function that best describes developmental change in processing speed during childhood and adolescence. In one sample, children and adolescents (N 5503) were tested twice over an average interval of 2 years on two psychometric measures of processing speed: Visual Matching and Cross Out. In another sample, children and adolescents (N 5 277) were tested four times, every 6 months, on Cross Out. Age-related changes in performance on both tasks were examined using six longitudinal models representing different hypotheses of growth. Linear, hyperbolic, inverse regression, and transition models yielded relatively poor fit to the data; the fit of the exponential and quadratic models was substantially better. The heuristic value of these latter models is discussed.

Powered by ScribeFire.

Technorati Tags: psychology, educational psychology, school psychology, neuropsychology, developmental, intelligence, IQ, IQ tests, WJ-R, WJ III, CHC, CHC theory, Visual Matching, Cross Out