Developmental Theory

Cognitive development typically refers to age-related changes in knowledge and acts of knowing, such as perceiving, remembering, problem solving, reasoning, and understanding. The development of cognition is studied most frequently in infants, children, and adolescents, where changes often are relatively rapid and striking. Many researchers also study cognitive development in aging adults, in children and adults during recovery of function following brain damage, and in a variety of species other than humans. Since the 1890s, when researchers such as James Mark Baldwin and Alfred Binet established cognitive development as a substantive area of inquiry, two overlapping goals have been evident. One goal is to provide insights into how complex, organized knowledge systems develop, an issue with a long history in philosophy and science. The other goal is to provide insights into optimizing human development, especially with respect to education. Researchers have adopted many different theoretical approaches to the study of cognitive development over the past 100 years, and they continue to do so. During the latter part of the twentieth century a relatively new approach, information processing, gained a degree of ascendancy because of its potential for providing rich insights into how cognition develops and how instruction might be improved.

Assumptions and Findings

In the 1950s and 1960s researchers began to notice similarities between human thinking and the new computers of that era, which could manipulate not only numbers but also a variety of nonnumeric symbols. Allen Newell and Herbert Simon were among the first to suggest that humans and computers could both be viewed as general symbol manipulators, and that knowledge of computers could be used as a metaphor for exploring human cognition. The argument that emerged was not that humans are computers, but rather that computers could be used as a source of ideas about how human cognition works and also as a tool for expressing ideas about how humans process information mentally. Information-processing studies of cognition and its development began to flourish in the 1960s and 1970s.

Information processing is not a theory of cognition but rather a general framework that comprises a family of theories sharing certain core assumptions. One assumption is that all cognitive activities involve mental processes that operate over real time on internal, symbolic representations of information. That is, information of all sorts–including the words on this page, memories of past events, knowledge about friends or world events, and abstract concepts such as "justice"–are all coded as mental representations with certain structural properties.

When one sees a painting for the first time, for example, perceptual processes code new sensory information and may also create more elaborate representations of what is seen. Memory processes store these representations and also retrieve previously developed representations that can be useful for interpreting the painting. Problem-solving and reasoning processes operate to help understand the artist's intent in creating the painting. From an information-processing view, one does not simply experience the painting. Instead, one is engaged in a series of events in which mental representations are created and manipulated by processes operating over time. Information-processing researchers seek to identify these processes and representations and to understand their properties. Researchers therefore focus less on whether children solve problems correctly and moreon how problems are solved. This approach has led to a rich set of findings about the skills and knowledge children acquire on specific tasks in such domains as reading, mathematics, and scientific thinking.

A second assumption is that these processes and representations exist within an organized system with definable properties and constraints. An important goal of research is to define the cognitive architecture, that is, the general structural characteristics of the information-processing system. For example, the amount of information that can be activated at any one time is limited, as is often evident when people try to remember new telephone numbers or solve difficult problems. This phenomenon is often interpreted in terms of working memory, an important, limited capacity system for manipulating information. Research on working memory has revealed the operation of three interacting components: a phonological loop for storing speech-basedinformation; a visual-spatial sketchpad for storing visual information; and an executive system for combining information from various sources to solve problems and create plans. New research, such as that reported in 2000 by Susan E. Gathercole and Susan J. Pickering, is beginning to link developmental change and individual differences in cognitive performance to changes in these components of working memory. Another constraint on cognitive processing is the speed at which processes operate. In general, faster processing speed should enable more competent performance on particular tasks. Not only does general processing speed increase from early childhood through adolescence, but as researcher Robert Kail reported in 1991, it does so at a consistent and well-defined rate of change. The reasons for this phenomenon still are not understood.

A third assumption is that cognitive development occurs via self-modification of the information-processing system. Although environmental events critically influence development, the mechanisms by which the information-processing system changes over time are assumed to be internal to the system itself. A number of such mechanisms have been proposed. For example, as children develop some processes become automatized in the sense that they are executed more rapidly and with less demand on limited attentional capacity than earlier in development. According to some theories, increasing automatization allows children to operate at higher levels of complexity and flexibility. Knowledge modification processes, such as generalization and discrimination, operate to create more powerful and accurate processes and representations. A critical task for developmental theorists is defining a cognitive architecture and self-modification mechanisms that, together, can account for the striking changes in thinking that emerge as children develop.

Information-processing theories of development differ significantly from other approaches in fundamental ways. They are not phenomenological because they are not limited to conscious experience, and they are not neurological in that they do not rely on neural or biochemical mechanisms as explanations. They differ from traditional stimulus-responses theories because of their emphasis on detailed descriptions of mental processes and representations that interact over time. Unlike structural theories, such as that of Jean Piaget, the focus is on very specific processes and representations that underlie performance. Information-processing theories often can be amalgamated to some extent with these other approaches, however. In contemporary neuroscience research, for example, information-processing concepts, such as working memory and processing speed, are often used to explore relations between brain and behavior.

Methods

The assumptions of information processing have led researchers to adapt or create methods appropriate for identifying processes, representations, and characteristics of cognitive architecture. Given the emphasis on temporal properties of processes, researchers have developed highly specialized, chronometric methods for measuring the speed of particular mental processes. With rule assessment, tasks are structured so that patterns of responses can be used to identify particular processes and decision rules. Protocol analysis is used to examine verbal self-reports, provided by participants as they solve problems, for evidence about solution procedures, internal representations, and processing constraints. When applied to the study of development, these methods need to be used carefully so that they are equally sensitive to important aspects of performance at different developmental levels.

Information-processing researchers also have adopted a number of distinctive methods for illustrating or representing their theories. Because of the emphasis on specific processes and their organization, flow charts and diagrams often are used to indicate how processing is structured. Some researchers take a more formal approach: They implement their theories of cognitive development as computer programs. To the extent that the programs mimic children's behavior and development, researchers receive some support for the veridicality of their theory. If, however, the program crashes, then clearly the theory is lacking.

Educational Implications

Ideally, educational assessment would provide specific insights about how to adapt instruction to individual children so as to optimize learning. In principle, information processing should provide a basis for assessing specific strengths and weaknesses and for identifying specific processes and representations that can be targeted for instruction. Teachers want their students to answer problems correctly, but measuring achievement only in terms of correct answers can be misleading: Often children can answer a problem correctly but for the wrong reasons, or incorrectly but for reasons that make sense. More important than answering correctly, in terms of educational goals, are whether students use appropriate solution strategies and whether they understand what they are doing. The value of information-processing research for education lies in its inherent distinction between the products of children's thinking (i.e., whether children solve problems correctly) and the processes (i.e., how problems are solved). Research on the development of school-related knowledge and skills is beginning to yield impressive advances.

In studies of young children's arithmetic, for example, researchers have identified a wide range of solution procedures, correct and incorrect, that children use to solve problems. To account for how children select among these procedures, how procedures change as children gain experience, and how some new procedures arise, Robert S. Siegler and Christopher Shipley (1995) developed an information-processing model that includes assumptions about an associative memory for number facts, a memory system for recording the results of past solutions, and a system for deciding whether and how to apply particular procedures. This model accounts extremely well for some aspects of children's development in arithmetic, and it has some specific instructional implications. For example, according to this model, associating problems and correct solutions is critical for later development of efficient solution procedures. Discouraging children from counting accurately with their fingers may increase the chance of incorrect associations developing and thus delay the use of more advanced procedures. The model is far from complete, but it provides a coherent basis for analyzing how children solve arithmetic problems, how and why change occurs, and how instruction might be adapted to the needs of individual children.

Similar progress has been made in other areas. Reading, for example, is a complex skill consisting of numerous components, and information-processing methods have been useful for identifying and measuring these components. One such component is phonological awareness, which includes the ability to identify and manipulate phonemes. Lynette Bradley and Peter E. Bryant (1983) found that instruction designed to enhance phonological awareness in young children strongly and positively influences the rate at which they become effective readers. Problem solving is critical to success in many academic domains. Amarjit S. Dhillon (1998) studied the behavior of experts and novices as they solved physics problems and found that their strategies could be analyzed in terms of fourteen processes or activities. Experts and novices differed systematically in the use and sequence of these activities, a finding that provides insights into understanding students' knowledge in terms of specific concepts and procedures. The results of this research were used to develop problem-solving instruction for high school and university students.

Aside from its use in specific academic domains, information processing also has provided a basis for assessing broad intellectual skills. A new generation of tests is emerging that are constructed so that children's performance can be interpreted in terms of relatively specific processing skills that, in principle, may be amenable to targeted instruction. One example is the Cognitive Assessment System (CAS), developed by Jagannath P. Das and Jack A. Naglieri, in which tasks from information-processing research have been adapted to measure four aspects of processing (planning, attention, simultaneous processing, and successive processing) that are emphasized in a comprehensive theory developed by the neuropsychologist Aleksandr Luria. Because of the links between theory and measures, the CAS has proved useful in interpreting performance for children with or without learning disabilities and for developing specific instructional interventions.

Prospects

Information processing is by no means the only approach for studying cognitive development, but its assumptions and methods have proved helpful in exploring the many ways in which children's thinking changes with development. Its greatest utility to date has been in studying task-specific or domain-specific processes and representations. It has been applied with somewhat less success to domain-general characteristics of development, as well as to topics such as motivation and affect that are critical to understanding development and optimizing education. At this point, it is not clear whether these apparent deficiencies are inherent to information processing or whether they are simply a result of how information-processing concepts and methods have been applied to date. The information-processing approach is challenged by connectionist and dynamic systems theories that do not share the assumptions about symbolic representations and discrete processes; by ecological theories that focus on environmental factors and their structure; by neuroscientific theories that provide explanations in terms of neural functioning and neuroanatomy; and by traditional theories, such as those of Jean Piaget and Lev Vygotsky, in which a more general level of analysis and explanation is emphasized. The extent to which information processing succeeds will depend, in part, on the extent to which its practitioners can adapt to accommodate these challenges and contribute to research that enriches educational assessment and instruction.