Unities in inductive reasoning.

Abstract

Three experiments sought to discover sources of communalities in performance on three inductive reasoning tasks: analogies, series completions, and classifications. In Experiment 1, 30 subjects completed an untuned pencil-and-paper test in Which they were asked to solve 90 induction items with names of mammals as content. Items were equally divided among the three kinds of induction tasks noted above. The subjects' task was to rank order four response options in terms of their goodness of fit as completions for each particular item. Data sets for the three tasks were highly intercorrelated, and a single exponential model of response choice provided a good fit to each data set. In Experiment 2,36 subjects completed a timed test in which they, too, were asked to solve 90 mammal-name induction items. Items were again equally divided among the three kinds of induction tasks noted above. The subjects' task was to choose the better of two response options as a completion for each particular item. Data sets for the three tasks were again highly intercorrelated, and a single linear model of response times provided a good fit to each data set. In Experiment 3, 18 subjects were timed while solving 1,440 induction items with schematic picture, verbal, and geometric content. Items were apprpximately equally divided among the three kinds of induction tasks noted above. Both analysis of stimulus and of subject variance supported the notion of highly related performance algorithms on the three tasks. It was concluded that a common or highly similar model of response choice and of information processing can account for at least some of the previously observed relationships in performance across induction tasks. The implications of these findings for psychometric and information-processing accounts of induction and intelligence are discussed.

A cognitive analysis of number-series problems: Sources of individual differences in performance

“…The differences in recognition of memorized sequences and calculation between high-and low-skill individuals are consistent with two recent observations about performance on various inductive reasoning tasks: (1) estimates of various information-processing parameters correlate significantly with factor loadings that represent psychometric measures of inductive reasoning (Sternberg & Gardner, 1983) and (2) adults who score well on tests of inductive reasoning generally process information more efficiently than other adults (Goldman & Pellegrino, 1984).…”

“…This approach is quite limited in scope, but it does have some advantages. Consider, for example, the research by Sternberg and Gardner (1983) on inductive reasoning. Sternberg and Gardner postulated such components as inference, mapping, and justification to account for performance on series-completion problems and on other tasks of inductive reasoning.…”

Cognitive, Metacognitive, and Motivational Aspects of Problem Solving

“…This instructional episode is based on a componential analysis of analogical reasoning (Sternberg, 1985;Sternberg & Gardner, 1983). In componential analysis, a reasoning task is broken down into its constituent cognitive processes.…”

A componential model for mental addition.

“…Componential models have been proposed for several types of reasoning tasks (e.g., Sternberg, 1977;Sternberg & Gardner, 1983), for verbal ability (e.g., Hunt, 1978;Hunt, Lunneborg, & Lewis, 1975), and for spatial ability (e.g., Pellegrino & Kail, 1982), among others. In the present study, the application of componential analysis is extended to numerical facility tasks through intensive and extensive validation of elementary information processes underlying performance on four types of mental addition problem.…”