Tuesday, March 22, 2005

Carroll Analyses of 50 CHC tests: Rnd 2: Gf and the executive controlled model of working memory

The Gf findings
One of the most interesting and unexpected findings to emerge from the Carroll Analyses of 50 CHC-designed tests (see 3-17 & 3-18-05 posts) is the composition of the Gf (fluid reasoning) factor. The presence of tests of inductive (Concept Formation) and deductive reasoning (Analysis-Synthesis) are consistent with decades of research—induction and deductive reasoning are considered the hallmarks of Gf. But, how does one explain the significant loadings for a test that requires the following of increasing complex series of oral directions (WJ III Understanding Directions) and an oral auditory analysis test that requires individuals to apply different processes (rhyming, substitution, deletion and reversal- Sound Awareness)?

First, as way of additional background, I have found (yes, in a series of other unpublished “languishing or banned to forgotten sectors of my hard drive” analyses of the WJIII+DS tests) that the WJ III Understanding Directions and Sound Awareness tests are two of the:

  • (a) best WJ III test-level predictors of academic achievement across language arts (reading and writing) and math, (b) highest g-loading tests on the first extracted principal component, and (c) most cognitively complex tests as defined by Guttmann’s Radex model (using multidimensional scaling). Clearly the WJ III UD and SA awareness tests are measures of complex cognitive processes associated with higher level abstract reasoning (Gf).
The current hypothesis
It is my current working hypothesis that the WJ III UD and SA tests require much in the way of working memory and executive/controlled attention, abilities that have been repeatedly linked in contemporary research on working memory and Gf (particularly as articulated by Engle & Kane and colleagues and as recently summarized by Horn and Blankson [2005] in the CIA2 book.) This view states that Gf abilities involve the process of:

  • "(1) gaining awareness of information (attention) and (2) holding different aspects of information in the span of awareness (working memory), both of which are dependent on (3) a capacity for maintaining concentration" (Horn & Blankson, 2005, p. 55)

In other words, the capacity to focus and maintain attention is now being seen as a critical component for successful performance on tasks that require higher-level fluid reasoning (Gf). According to Horn and Blankson, “focusing and maintaining attention appears to be an aspect of the capacity for apprehension that Spearman described as a major feature of g” (p. 54; emphasis added by me).

From my readings, Engle and colleagues appear to have developed the most systematic program of research and theoretical explanation for the “controlled-attention” (or executive attention) view of working memory, it’s role in many real world phenomena, and more importantly, the hypothesis that “executive attentional abilities are some way related to general fluid intelligence” (Heitz, Unsworth & Engle, 2005, p. 74). Executive or controlled attention enhances performance in situations that require selective or controlled attention, the ability to switch between plans and strategies, and the inhibition of task-irrelevant irrelevant information (intrusions) in working memory, abilities that collectively contribute to the ability to solve complex reasoning problems in working memory (e.g., WJ III Concept Formation – a measure of Induction).

A caveat
Of course, the question is still not answered (from the current factor analysis based results) whether the executive/controlled attention component of working memory is a component of, or a causal mechanism, of Gf. The later causal view is just as probable. Empirical evidence for the causal hypothesis, based on these same data, can be viewed in a summary of the working memory/Gf causal research in my CIA2 chapter or by viewing a pre-publication version of that chapter (click here)

Concluding comments

The Gf factor reported in the Carroll Analyses of 50 CHC-tests is consistent with contemporary research that suggests a significant link between fluid reasoning (Gf) and the executive/controlled attention model of working memory.

From a practical perspective, these findings suggest that assessment professionals who are evaluating the ability of a person to perform complex cognitive reasoning and problem solving tasks (Gf), if using the WJ III battery, should attend to (aside from the WJ III Gf tests) an individuals performance on the WJ III Understanding Directions and Sound Awareness tests (and possibly WJ III Story Recall as well). Performance on these tests may provide insights on a person's ability to sustain focused executive control during the activite manipulation of multidimensional pieces of information in working memory--an ability that may be important when tasks require complex fluid reasoning ability (Gf).

5 comments:

Matthew Reynolds said...

I think that these are the same

Matthew Reynolds said...

Have to finish my thought...a non-WJ test that assessment professionals might want to consider is Arithmetic from the WISC. Arithmetic loads on Gf and also has the highest g-loading of WISC subtests. It seems as if this subtest also requires executive/controlled attention.

John Garruto said...

I've spoken about this some time before but wanted to mention it again.

I have always found the Boolean connection between Concept Formation and Understanding Directions to be an intriguing one. The later portions of CF require an understanding of some of the basic principles of logic. Similarly, understanding directions seems to simultaneously tap working memory (holding the information) and the Boolean connection (i.e. "If the lights are on and the dog is white, point to the bed").

It's pretty much the converse of some of the Ga. Basically Ga you either pretty much have or you don't (hence needing to get most of the items on sound awareness right for an average score.) Conversely, it's rare that I see elementary students make it past the Boolean logic on CF and when they DO...I can usually say with some degree of conviction that inductive reasoning processes are stable.

Kevin said...

John. I've always liked your hypothesis about the shared Boolean connetion between Concept Formation and Understanding Directions. And, as per the executive/controlled model of working memory, being able to maintain all the different "parts" of each Boolean set, in order to then generate a solution or respond, makes sense. One cannot complete a sequence of Boolean logic if pieces of the logic can't be actively maintained in working memory. Good stuff.

Teddy Croston said...

I do not think so.