What one intelligence test measures: A theoretical account of the processing in the Raven Progressive Matrices Test.

Abstract

The cognitive processes in a widely used, nonverbal test of analytic intelligence, the Raven Progressive Matrices Test (Raven, 1962), are analyzed in terms of which processes distinguish between higher scoring and lower scoring subjects and which processes are common to all subjects and all items on the test. The analysis is based on detailed performance characteristics, such as verbal protocols, eye-fixation patterns, and errors. The theory is expressed as a pair of computer simulation models that perform like the median or best college students in the sample. The processing characteristic common to all subjects is an incremental, reiterative strategy for encoding and inducing the regularities in each problem. The processes that distinguish among individuals are primarily the ability to induce abstract relations and the ability to dynamically manage a large set of problemsolving goals in working memory.In this article, we analyze a form of thinking that is prototypical of what psychologists consider to be analytic intelligence. We use the term analytic intelligence to refer to the ability to reason and solve problems involving new information, without relying extensively on an explicit base of declarative knowledge derived from either schooling or previous experience. In the theory of R. Cattell (1963), this form of intelligence has been labeled fluid intelligence and has been contrasted with crystallized intelligence, which more directly reflects the previously acquired knowledge and skills that have been crystallized with experience. Thus, analytic intelligence refers to the ability to deal with novelty, to adapt one's thinking to a new cognitive problem. In this article, we provide a theoretical account of what it means to perform well on a classic test of analytic intelligence, the Raven Progressive Matrices Test (Raven, 1962).We describe a detailed theoretical model of the processes used in solving the Raven test, contrasting the performance of college students who are less successful in solving the problems with those who are more successful. The model is based on multiple dependent measures, including verbal reports, eye fixations, and patterns of errors on different types of problems. The experimental investigations led to the development of computer simulation models that test the sufficiency of our analysis. Two This work was supported in part by Contract N00014-89-J-1218 from the Office of Naval Research and by Research Scientist Development Awards MH-00661 and MH-00662 from the National Institute of Mental Health. The simulation models were programmed by Peter Shell and Craig Bearer.We are grateful to Earl Hunt, David Klahr, Kenneth Kotovsky, Allen Newell, James Pellegrino, Herbert Simon, and Robert Sternberg for their constructive comments on drafts of the manuscript. We also want to acknowledge the help of David Fallside in implementing and analyzing the Tower of Hanoi experiment.Correspondence concerning this article should be addressed to Patricia A. Carpenter, Department of Psychology, Carnegie...

Eye movements in reading and information processing: 20 years of research.

“…Eye movements have also been recorded as participants solve math and physics problems (De Corte, Verschaffel, & Pauwels, 1990;Hegarty & Just, 1993;Hegarty, Mayer, & Green, 1992;Suppes, 1990;Suppes, Cohen, Laddaga, Anliker, & Floyd, 1983;Verschaffel, De Corte, & Pauwels, 1992), as well as analogies (Bethell-Fox, Lohman, & Snow, 1984;Carpenter, Just, & Shell, 1990;Dillon, 1985). Not surprisingly, more complicated aspects of the problems typically lead to more and longer fixations.…”

A capacity theory of comprehension: Individual differences in working memory.

“…Some of these implications have recently been confirmed in the area of complex problem solving, specifically in solving the problems in the Raven Progressive Matrices Test. Subjects with lower scores in the Raven test were less able to store intermediate goals, as indicated in an independent task, and as indicated by their disadvantage on Raven test items that required the storage of a large number of subgoals and partial solutions (Carpenter, Just, & Shell, 1990).…”

The role of prefrontal cortex in working-memory capacity, executive attention, and general fluid intelligence: An individual-differences perspective

“…Thus, when relations among stimuli demanded integration-perhaps accomplished through WM maintenance-PFC patients were markedly impaired in a Gf-loaded task of induction. Indeed, a computational model of performance on such matrix-reasoning tasks indicates that integration of multiple stimulus dimensions relies on WM capacity (Carpenter, Just, & Shell, 1990).…”

“…More recently, an f MRI study examined cortical activation patterns during the performance of different subtypes of Raven's problems (Prabhakaran, Smith, Desmond, Glover, & Gabrieli, 1997). Some problems required only visuospatial analysis, and others additionally required nonspatial, abstract/analytical reasoning (see Carpenter et al, 1990). Relative to a no-delay, matching-to-sample baseline, spatial reasoning activated right dorsolateral areas (Brodmann 9,46) along with other right hemisphere regions such as anterior cingulate (32), parietal areas (7,40), and temporal areas (37).…”