Fluid reasoning (Gf) performance and response generation speed

During the past few months this blogmaster has provided a number of empirically-based insights regarding individual differences in fluid reasoning (Gf) performance.

Below is yet another potentialy usefull Gf interpretation tidbit.

Verguts, T. & De Boeck, P. (2000). Generation Speed in Raven's Progressive Matrices Test. Intelligence, 27(4), 329-345.

• Performance on fluid reasoning (Gf) tests (e.g., Ravens Progressive Matrices-RPM) may be enhanced by the speed/fluency by which individuals identify rules that govern Gf test items. When faced with fluid reasoning tasks, individuals are viewed to have, at their disposal, a pool or distribution of rules from which to select. The fluency by which an individual “samples” or generates rules (response generation speed) was statistically linked to Gf performance in Verguts and De Boeck’s (2000) study in a sample of 127 undergraduate students.

• This finding is not new. As early as 1898 (Thorndike) noted that in order to generate correct responses to problems, an individual must first generate a number of possibilities, retain them, and then implement the correct possibility/rule. Verbal response fluency has been studied extensively (see Carroll, 1993 for an overview) while Gf-related fluency has not.

• Verguts and DeBoeck (2000) suggest that “If…rules are compared with balls in an urn, this means that people sample balls from an urn. Individual differences in the generation process can be thought of as sampling from different urns (qualitative differences) or at different rate (quantitative differences)” (p.330).

• These investigators demonstrated that response or rule generation speed was correlated with Gf performance (viz., RPM test performance), particularly on items/tasks where discovering the rule(s) is more challenging. However, speed of rule generation should be considered a necessary, but not sufficient condition, for optimal Gf performance. According to these investigators, other variables that may influence rule generation fluency/speed may include individual differences in (a) the quality of rules sampled and (b) the accuracy of applying the generated results (which may be related to working memory efficiency).