Intelligence and Memory

views updated


Memory plays an important role in intelligent behavior. Modern assessments of cognitive functioning, such as the Wechsler Adult Intelligence Scale (WAIS III), include memory tests among other essential measure of mental capability. Modern information-processing conceptions of memory, however, represent a distinct change from the information-reservoir conceptions of memory put forth by earlier researchers such as Richard Atkinson and Richard Shiffrin (1968). This change, a consequence of the shift from using storage metaphors for memory to using computer processing metaphors, has revolutionized scientific understanding of memory and its link to intelligence.

Memory as Information Processing

In their edited volume, Akira Miyake and Priti Shah (1999) integrated the work of leading memory theorists into a consensus definition of memory, which dispenses with the notion of a biochemical filing cabinet in the brain and proposes a multifaceted, working information-processing system whose function is to aid in complex cognition. As illustrated in Figure 1, the limitations of this information-processing system can be attributed to multiple factors: such as the encoding of novel information, the simultaneous storage and processing of information, and the retrieval of information stored long term. The concept of working memory is not new but was first put forth by Alan Baddeley and Graham Hitch in 1974. However, much hard work since the 1970s has extended Baddeley and Hitch's initial work to develop a more precise characterization of this system, its limitations, and its relationship to intelligent behavior.

Measures of working memory have shown a strong and consistent relationship with measures of intelligence and complex cognitive processing. Though numerous experimental findings demonstrate this link, those of Patrick Kyllonen and his colleagues, Raymond Christal and Deborah Stephens, have been especially illuminating. Seeking to provide an information-processing explanation for intelligence, Kyllonen and his colleagues developed and tested his model of intelligent performance, which posited that mental functioning depends on four cognitive sources: working memory, processing speed, declarative knowledge, and procedural knowledge. They found that of the four sources, working memory showed the strongest relations to skill acquisition and performance on intelligence tests (Kyllonen, 1996; Kyllonen and Stephens, 1990). Kyllonen discovered a high correlation between measures of working memory and tests of reasoning, even concluding that working memory and general intelligence amount to the same thing (Kyllonen, 2002).

Other work has been geared toward determining more precisely what accounts for the relationship between working memory and cognitive functioning. Researchers have focused on the encoding, processing, and retrieval aspects of the working memory system.

The Role of Encoding

Encoding involves transforming perceptual information from the environment into initial input for the information-processing system (see Figure 1). Limitations in receiving information from the environment have implications for later information processing and, ultimately, intelligent behavior.

Christopher Jarrold, Alan Baddeley, and Alexa Hewes (2000) examined the relative amount of immediate recall, near-immediate recall, and short-term recall in children of varying levels of retardation. They briefly presented the children with three spoken words, each of which was associated with one of three horizontal positions on a computer screen such that the first word spoken was associated with the leftmost position. They then highlighted one of the three positions and asked the children to recall the word that had been associated with that position. Recalling the word associated with the leftmost position, spoken earliest, required short-term recall, the middle position, near-immediate recall, and the rightmost position, immediate recall.

These researchers were interested in determining whether failure to rehearse to-be-remembered information could account for memory deficits in children with Down's syndrome. They discovered, however, that differences in short-term recall associated with intellectual ability occurred despite the fact that none of the children used rehearsal as a memory aid. Further, these memory differences occurred only during short-term recall, in which children with Down's syndrome demonstrated memory deficits that the other children did not; all of the children had equally poor near-immediate recall and equally good immediate recall. Jarrold, Baddeley, and Hewes suggested that one possible explanation for these results could be encoding limitation—even though their perceptual processing was equivalent, children with Down's syndrome could not transform environmental information into system input (i.e., memory traces) as efficiently as the other children with lesser intelligence deficits could.

This conclusion mirrors that of Norman Ellis and Darlene Meador (1985), who investigated mnemonic strategies and short-term recall deficits in retarded children. They presented children with an experimental stimulus and then, after a delay, presented a probe stimulus. They asked the children to compare the probe stimulus with their memory of the experimental stimulus and indicate whether they matched. As expected, recognition accuracy on this task decreased as IQ decreased. More important, rates of forgetting across different levels of IQ were equivalent on this task, even with the precluding of mnemonic strategies. Performance differences correlated with IQ differences with the simultaneous presentation of experimental and probe stimuli, even at retention intervals of twenty seconds.

The Role of Processing

As illustrated in Figure 1, after encoding, information must remain temporarily active during its support of cognitive activity. Limitations in simultaneously storing and processing information have implications for complex cognition because the active maintenance and processing of information in working memory plays an important role in intelligent behavior. There are, however, differing views on the nature of limitations in this aspect of the working-memory system, whether it is task-specific or task-independent.

Meredyth Daneman and her colleagues, Patricia Carpenter and Brenda Hannon, have tested the hypothesis that the efficiency of task-specific processing is the critical link between working memory and complex cognition. This work and related research helped to develop several working-memory tests, each with a storage and processing aspect (see Figure 2). The critical difference between these tests is the type of processing required in each. The reading-span test requires verbal processing specific to reading in addition to verbal storage. The individual must read a set of sentences and, after finishing reading all of the sentences in the set, recall the last word of each sentence. The speaking-span test is similar in calling for the remembering of a set of target words but calls for the presentation of the target words first; the individual must remember the words while using each of them to orally generate a sentence, a task that requires verbal fluency. The operation-span test, created by Marilyn Turner and Randall Engle (1989), requires mathematical processing: After completing a set of simple mathematical operations, the individual must recall the word associated with each one.

Findings from research using these memory tests suggest that the correlation between memory and task performance depends on the degree of similarity between the type of processing required by the task and the memory test. Daneman (1991) found, for example, that speaking-span scores showed a stronger relation to a measure of oral fluency than did reading-span scores. Conversely, reading-span scores showed a stronger relation to oral reading skills than did speaking-span scores. Daneman and Hannon (2001) demonstrated that operation-span scores showed a relatively weaker correlation to reading comprehension than did reading-span scores.

Not all research agrees with the conclusion the processes on memory tests must be task-specific. In contrast, Engle his colleagues have repeatedly demonstrated that the relationship between memory and cognition is not mediated by efficient task-specific processing. Engle, Kane, and Tuholski (1999) described a series of studies in which task-specific processing efficiency was statistically or experimentally controlled but in which the relationship between working memory and performance on various tests was not eliminated. In one study, the mathematical processing demands of the operation span test were equated across individuals. The mathematical ability of each participant was determined, and the difficulty of the operations administered in the operation span test was tailored to ability level. Controlling for individual differences in mathematical ability failed to reduce the strong correlation between performance on the memory test and reading comprehension.

The operation span test required the allocation of attention to both completing mathematical operations and remembering target words, but equating mathematical ability did not reduce the correlation between working memory and complex cognition. Engle and his colleagues argued, therefore, that domain-free controlled attention is the essence of working-memory limitations and drives the relationship between measures of working memory and measures of intelligence. They claimed that controlled attention activates information, either from the immediate environment or from long-term memory, and maintains that information in memory while it is being processed, particularly in the face of distraction. Tuholski, Engle, and Baylis (2001) found that performance on the operation span test could predict distractibility on a computerized counting task, indicating the role of controlled attention in reducing the impact of distraction.

Regardless of disagreement over the nature of working-memory limitations, the work of both Daneman and her colleagues and Engle and his colleagues clearly indicates that the storage aspect of memory alone does not account for its strong relationship to complex cognition. Daneman and Carpenter (1980) found that performance on the reading span test correlated substantially with verbal Scholastic Assessment Test (SAT) scores and reading comprehension, whereas a test of verbal storage—memorizing a list of words—showed only moderate correlations with SAT scores and reading comprehension. Consistent with these findings, Engle and his colleagues (1999) used structural equation modeling techniques to demonstrate that working memory had a stronger correlation with general reasoning than did short-term storage. They also showed that working-memory measures were better predictors of verbal and quantitative SAT scores than was short-term storage.

The Role of Knowledge and Skills

An emerging interest in understanding everyday cognitive functioning has turned the focus of some researchers to the role of long-term knowledge and skills in the memory system. These researchers are wary of experimental results based on performance on simple laboratory tasks such as simple mental calculations. They argue that, in typical college student samples, such tasks highlight working memory demands at the expense of acquired knowledge and experience, which is a critical aspect of everyday functioning. Instead, they study how highly developed skills and knowledge influence memory performance.

Anders Ericsson and Peter Delaney (1999) proposed that the functioning of the working-memory system during highly skilled performance is determined by specialized encoding and retrieval mechanisms formed over extended periods of practice. They claimed that individuals use these mechanisms to encode information from the environment into long-term memory to enable efficient and extensive retrieval of this information later on. These mechanisms allow the individual to demonstrate an unexpectedly large skill-specific memory capacity because of the ready availability of information relevant to that skill that readily available in long-term memory.

Ericsson and Delaney found support for their theory in the verbalizations of memory experts, people who, through training, have developed above-average memory skills. These verbal reports indicated that memory experts use existing knowledge to encode information into larger chunks for later retrieval. For example, to memorize a very large string of digits, one expert reported encoding the digits as mile running times and retrieving the digit string as groups of digits rather than as individual numbers. When digits were presented in such a way that they could not be associated with existing knowledge (e.g., they would make nonsensical running times), there was no expansion of memory. In addition, Ericsson and Delaney reported that chess experts not only encoded individual chess pieces according to meaningful configurations but also developed retrieval structures associated with locations on the chessboard.

Although memory assessment is still a critical aspect of commercial intelligence tests, such as the WAIS-III and its companion Wechsler Memory Scale (WMS), memory testing has changed. In contrast to the WAIS-R, the WAIS-III no longer features digit span as a mandatory test, but it does feature several additional tests of working memory. Also consistent with the findings of recent research, the WMS features tests of immediate, delayed, and working memory. Instead of reflecting a static storage capacity, the memory aspect of modern intelligence tests now reflects the dynamic functioning of a complex, multifaceted information-processing system. Such changes in memory and intelligence testing indicate growing understanding of the critical role of information processing in complex cognition.


Atkinson, R. C., and Shiffrin, R. M. (1968). Working memory. In G. H. Bower, ed., The psychology of learning and motivation: Advances in research and theory, Vol 2. New York: Academic Press.

Baddeley, A. D., and Carpenter, P. A. (1980). Individual differences in working memory and reading. Journal of Verbal Learning and Verbal Behavior 19, 450-466.

Baddeley, A. D., and Hannon, B. (2001). Using working memory theory to investigate the construct validity of multiple-choice reading comprehension tests such as the SAT. Journal of Experimental Psychology: General 130, 208-223.

Baddeley, A. D., and Hitch, G. (1974). Working memory. In G. H. Bower, ed., The psychology of learning and motivation: Advances in research and theory, Vol. 2. New York: Academic Press.

Baddeley, A. D., and Logie, R. H. (1999). Working memory: The multiple-component model. In A. Miyake and P. Shah, eds., Models of working memory: Mechanisms of active maintenance and executive control. Cambridge, UK: Cambridge University Press.

Daneman, M. (1991). Working memory as a predictor of verbal fluency. Journal of Psycholinguistic Research 20, 445-464.

Ellis, N., and Meador, D. M. (1985). Forgetting in retarded and nonretarded persons under conditions of minimal strategy use. Intelligence 9, 87-96.

Engle, R. W., Kane, M. J., and Tuholski, S. W. (1999). Individual differences in working memory capacity and what they tell us about controlled attention, general fluid intelligence, and functions of the prefrontal cortex. In A. Miyake and P. Shah, eds., Models of working memory: Mechanisms of active maintenance and executive control. Cambridge, UK: Cambridge University Press.

Engle, R. W., Tuholski, S. W., Laughlin, J. E., and Conway, A. R. A. (1999). Working memory, short-term memory and general fluid intelligence: A latent variable approach. Journal of Experimental Psychology: General 128, 309-331.

Ericsson, K. A., and Delaney, P. F. (1999). Long-term working memory as an alternative to capacity models of working memory in everyday skilled performance. In A. Miyake and P. Shah, eds., Models of working memory: Mechanisms of active maintenance and executive control. Cambridge, UK: Cambridge University Press.

Jarrold, C., Baddeley, A. D., and Hewes, A. K. (2000). Verbal short-term memory deficits in Down Syndrome: A consequence of problems in rehearsal? Journal of Child Psychology and Psychiatry and Allied Disciplines 41, 223-244.

Kyllonen, P. C. (1996). Is working memory capacity Spearman's g? In I. Dennis and P. Tapsfield, eds., Human abilities: Their nature and measurement. Mahwah, NJ: Erlbaum.

—— (2002). 'g:' knowledge, speed, strategies, or working- memory capacity? A systems perspective. In R. J. Sternberg and E. L. Grigorenko, eds., The general factor of intelligence: How general is it? Mahwah, NJ: Erlbaum.

Kyllonen, P. C., and Stephens, D. L. (1990). Cognitive abilities as determinants of success in acquiring logic skill. Learning and Individual Differences 2, 29-160.

Miyake, A., and Shah, P. (1999). Toward unified theories of working memory. In A. Miyake and P. Shah, eds., Models of working memory: Mechanisms of active maintenance and executive control. Cambridge, UK: Cambridge University Press.

Tuholski, S. W., Engle, R. W., and Baylis, G. C. (2001). Individual differences in working memory capacity and enumeration. Memory & Cognition 29, 484-492.

Turner, M. L., and Engle, R. W. (1989). Is working memory capacity task dependent? Journal of Memory and Language 28, 127-154.

Anna T.Cianciolo

Robert J.Sternberg

About this article

Intelligence and Memory

Updated About content Print Article


Intelligence and Memory