Cognitive Processes: Historical Perspective

Introduction

Cognitive processes range from the most basic (such as simple reaction time, choice reaction time, letter comparisons, and so on) to the most complex forms of human cognition (such as planning, attention, reasoning, and memory). Numerous attempts have been made to operationalize measurement of cognitive processes – indeed, process measures are prominent among the individual scales of most omnibus intellectual ability tests. However, the identification of individual differences in specific cognitive processes has been fraught with difficulties, due to two major factors. The first factor is that tests of cognitive processes tend to be correlated with one another. The second factor is that the content of the test items also determines individual differences in test performance, sometimes to a much greater degree than the underlying processes. The history of cognitive process assessment is described, and a brief review of contemporary issues and problems is presented.

In the hundred or so years of modern psychological assessment, there has been substantial interest in the efficient, reliable, and valid assessment of cognitive processes. The list of cognitive processes considered for assessment range from the most basic sensory and perceptual activities (such as brightness discrimination, and differential weight judgements) through to the most complex activities (such as analogical reasoning and creativity). A comprehensive list of cognitive processes studied through individual-differences assessments would include nearly all of the tasks studied by experimental psychologists concerned with discovering fundamental building blocks for mental life, and additional processes that are mostly of interest to differential psychologists.

Early Assessments of Cognitive Processes

In a classic Manual of Mental and Physical Tests, Whipple (1910/1914) divided the range of mental assessments into two broad categories, simpler processes and complex processes. For the simpler tests, Whipple listed sensory tests (e.g. colour blindness, discrimination of pitch, discrimination of lifted weights), and tests of attention and perception. Some of these tests were apparatus tests, while others were the kinds of paper and pencil tests familiar to modern psychologists. Measurement of ‘visual apprehension’ – that is, how many objects can be perceived in a brief presentation was administered with a tachistoscope, an [Page 242]instrument that could be adjusted to provide only the briefest exposure of the stimuli to the examinee. Tests of attention included cancellation tests (of which the modern Symbol-Digit test is an exemplar), simultaneous adding, and counting dots. Complex tests described by Whipple included tests of description, association, suggestibility, imagination, and intellectual ability. While one might argue that intellectual ability is not a cognitive ‘process’ per se, Binet's method for assessment of intellect was specifically predicated on an amalgamation of several different cognitive processes, such as recognition and recall forms of memory, visual and tactile judgements, and spatial visualization, among other processes (e.g. see Binet-Simon, 1905/1973). Even though the Binet-Simon scales are themselves measures of cognitive processes, people traditionally think of the Binet-Simon and more recent tests as intelligence tests, mainly because the Binet-Simon test yields a single amalgamated score (Mental Age), even though it is possible to examine the individual cognitive processes scale scores.

The Binet-Simon scales illustrate one of the most important characteristics of cognitive processes assessments. This characteristic is called ‘positive manifold’ – and it refers to the nearly universal property of mental assessments that they are positively intercorrelated. That is, in any large sample of examinees and cognitive process measures, an intercorrelation matrix of the measures will show positive correlations throughout the matrix. In simple terms, this means that all cognitive assessments tend to share some variance-individuals who perform well on one cognitive process assessment will also tend to perform better than average on another cognitive process assessment, even though the measures may seem to assess theoretically and practically different cognitive processes. This property of cognitive process assessments made it possible for Binet and Simon to develop a coherent and comprehensive assessment of intelligence. Because the individual cognitive process scales were themselves positively and often substantially correlated, aggregation of the separate measures resulted in a diminution of scale-specific variance contributions, and an accentuation of the general intellectual ability, common to the specific process measures. The result was a robust measure that could be well-replicated with a wide variety of instruments, as long as there was a broad sampling of the underlying cognitive processes assessed.

...