Saturday, November 12, 2005

Structure of mental speed (Gs/Gt) and causal effect on fluid reasoning (Gf)

Danthiir, V., Wilhelm, O., Schulze, R., & Roberts, R. D. (2005). Factor structure and validity of paper-and-pencil measures of mental speed: Evidence for a higher-order model? Intelligence, 33(5), 491-514.

An interesting article in Intelligence regarding the possible higher-order structure of mental speed (Gs/Gt) and it’s relation to fluid intelligence (Gf). The contemporary structural research I recently reviewed regarding Gs and Gt provides for a slightly different conclusion (click here for a summary).

The strong reported Gs-to-Gf causal path continues to support the notion that cognitive processing speed, and possibly the larger domain of cognitive efficiency (Gs + Working memory), is important for higher-order thinking. See prior causal research summary that suggests the Gs-to-Gf relation may be more complex than simple Gs-to-Gf.

  • This study explored the structure of elementary cognitive tasks (ECTs) and relations between the corresponding construct(s) with processing speed (Gs) and fluid intelligence (Gf). Participants (N = 321) completed 14 ECTs, 3 Gs, and 6 Gf marker tests, all administered in paper-and-pencil format to reduce potential confounds evident when tasks are presented using different media. Factor analysis of the ECTs resulted in a general mental speed factor, along with several task-class specific factors. General mental speed was indistinguishable from Gs and highly correlated with Gf. Significant correlations were also found between Gf and variance specific to task-class speed factors. The findings point to the non-unitary nature of mental speed and the potentially important role of specific speed factors for examining the relationship between speed and fluid intelligence.

Additional author comments from manuscript

  • The first stated aim of the current study was to investigate the structure of cognitive speed, given there is no agreement in the literature as to whether cognitive speed is a unitary or a non-unitary construct.
  • The current study does add support to the conceptualization of cognitive speed as a multi-dimensional construct.
  • In confirmatory modeling of these tasks, a higher-order model incorporating a general mental speed factor and group factors specific to the Switching, OMO, Substitution, and Hick task classes fit the data well.
  • Due to the preliminary and exploratory nature of this study, only some speculations regarding the substantive nature of the factors will be presented.
  • One possibility is that the task-class specific variance reflects only method variance, as opposed to the speed of carrying out a particular cognitive process or processes not captured by the general speed factor.
  • The second aim, as stated, was to assess the relation between general mental speed, as measured by the ECTs, and Gs. Of note regarding this aim is the very near perfect relationship found between the general mental speed factor of the ECTs and the Gs factor. As mentioned, in this study, Gs was assessed only by tests of Perceptual Speed, making it arguable as to whether this relationship holds in regard to the broad factor of Gs, or the narrower factor of Perceptual Speed.
  • However, the structure of Gs is not well established, and as previously noted, Perceptual Speed is also considered to be a non-unitary factor (e.g., Ackerman et al., 2002; Ackerman & Cianciolo, 2000; French et al., 1963).
  • The third aim of the current study was to examine the relationships between any mental speed factors which were identified and fluid intelligence. In the final structural model, the general mental speed factor was substantially correlated (r = 0.54) with Gf.
  • One consideration that must be emphasized is that the size of the relationship between the speed factors and Gf may be overestimated in these results, due to the fact that time limits were imposed on the tasks used to assess fluid intelligence.

No comments: