Wednesday, February 24, 2010

Beyond CHC: Pushing the edge of the CHC and WJ III envelope

This is a revision of a prior post (which I have now deleted).  I previously posted an online presentation (at SlideShare) summarizing a series of recent analysis I completed. The analysis are grounded in a variety of exploratory data analysis I  conducted over the past nine years, which when combined with recent theoretical literature, resulted in my attempt to search for possible modifications and extensions to CHC model of intelligence, as operationally defined by the WJ III test battery.

The original SlideShare  file (and associated PDF copies) did not include the alternative WJ III measurement model that was the basis of the alternative CHC theoretical models presented. The current version now includes the alternative WJ III measurement model.

The show can be viewed by clicking here [NOTE----SlideShare is experiencing problems today and I can't access the site to see if the new sets of slides uploaded correctly.  Be patient and come back and try again if you get an error message].   Alternatively, clicking here provides access to a PDF copy of the complete set of slides for immediate viewing or download.  Clicking here provides access to a "brief" copy of the key conclusion slides (in PDF format)

The underlying measurement model presented requires an understanding of the task demands and abilities measured by the various Woodcock Johnson III tests. For readers unfamiliar with the names, test descriptions, and abilities measured, it is recommended that you first become familiar with the tests by reviewing a document available here

Below are the key summary slides from the brief PDF file. I am assuming that the reader is familiar with the WJ III tests and can recognize the implications of some of the new alternative interpretations for select CHC factors and WJ III tests. I have inserted some background slides that should help readers understand the idea behind some of the proposed alternative CHC theory models.

In this post I want to focus on the WJ III measurement model suggested.  I  draw attention to two of the WJ measurement model slides (see below) as I believe they contain potentially important new insights into possible CHC factors and supplemental interpretation material for select WJ III tests.

The first measurement slide suggests that broad Gs (processing speed) may subsume a number of domain-specific processing speed abilities differentiated by stimulus content. I believe the four different speeded factors might best be interpreted as fluency factors as not all the test indicators are performed against a time constraint; but fluency and efficiency of performance may be the key to understanding the commonality of the tests loading on the domain specific speeded or fluency factors (e.g., the loading of the Calculation test on the obvious quantitative or numerical fluency factor). Of the four hypothesized cognitive speed/ fluency factors (numerical or quantitative fluency, reading and writing fluency, visual processing fluency, and fluency of Gc abilities), I am most intrigued by the latter. I hypothesize that the common feature of the Gs(Gc) factor tests is speed of lexical access (how quickly and efficiently one can access ones lexicon), which is similar to the concept of rapid automatic naming (RAN).

The next slide also suggests some very interesting hypotheses regarding the nature of Gf (fluid intelligence) and complex working memory. As can be seen in the slide, two different fluid intelligence factors were suggested. One clearly represents quantitative reasoning (RQ). The other, which I current label Gf* in the figure, is very intriguing. Ever since the WJ III was published I have run a variety of exploratory data analysis (some of which are summarized in the slide show available through the online PPT SlideShare show). A robust finding across all analyses has been the constant association of the Sound Awareness and Understanding Directions tests on a single Gf factor. The current model  suggests that these two tests, which require little in the way of inductive or deductive reasoning or "thinking", most likely tap complex language-based working memory ability. When combined with the other tests that load on Gf*, the hypothesis is presented that in addition to a quantitative reasoning Gf factor, there may be a second (and more broad) Gf factor that deals with cognitive complexity both in terms of inductive and deductive reasoning demands and heavy demands placed upon a person's complex working memory resources. This is an intriguing finding consistent with the extant research literature that consistently suggests a strong relation between fluid intelligence and complex working memory.

I could go on and on and on and speculate for days.  Instead, I would prefer that those interested in discussing these findings do so via the CHC listserv and/or join IQs Corner at Facebook (click on IQs Corner Facebook badge at top of this blog) and start a discussion thread.

The results, interpretations, and opinions presented in the above described material reflect the opinions and interpretations of Kevin McGrew, a co-author of the WJ III battery [conflict of information disclosure]. The information does not necessarily represent the opinions of the other WJ III co-authors or the publisher of the WJ III.

Technorati Tags: , , , , , , , , , , , , , , , , , , , ,

[Double click images to enlarge]


Technorati Tags: , , , , , , , , , , , , , , ,

No comments: