Richard E Mayer. 21st Century Education: A Reference Handbook. Editor: Thomas L Good. Sage Publication. 2008.
The primary goal of education is to help people learn. More specifically, the goal is to help people learn in ways that will allow them to use what they have learned in new situations—a process that can be called problem-solving transfer (Mayer & Wittrock, 2006). To accomplish this goal, it is useful for educators to have a clear understanding of how the human mind works.
This chapter explores the information processing view of learning, which currently offers the most comprehensive, best supported, and most widely accepted theory of how people learn (Bransford, Brown, & Cocking, 1999; Bruning, Schraw, Norby, & Ronning, 2004; Mayer, 2008). In this chapter, I summarize the main tenets of information processing theory; compare information processing theory with other views of learning; summarize the implications of information processing theory for learning, instructing, and assessing; summarize contributions of the information processing view in education including psychologies of subject matter, cognitive process instruction, and instructional design; and explore future directions for theories of learning in 21st-century education.
Humans are processors of information. This simple statement summarizes the essence of information processing theory. According to the information processing view of how the human mind works, people take in information from the outside world through their eyes and ears, construct an internal mental representation, apply cognitive processes that mentally manipulate the representation, and use their representations to plan and carry out actions. As you can see, two key elements in information processing theory are cognitive representations and cognitive processes. Information from the outside world is transformed in the learner’s cognitive system by cognitive representations; cognitive processes then performmental computations, or the systematic manipulation of the learner’s knowledge.
In short, the “information” part of information processing refers to the learner’s cognitive representations and the “processing” part of information processing refers to the learner’s cognitive processes. Human information processing involves building, manipulating, and using cognitive representations. According to the information processing view, learning involves cognitive processing aimed at building cognitive representations.
Research in cognitive science offers three important principles that should be part of any educationally relevant theory of how people learn (Mayer, 2001, 2005a):
- Dual channels: People have separate channels for processing verbal material and pictorial material (Paivio, 1986).
- Limited capacity: Within each channel people are able to attend to only a few pieces of information at any one time (Baddeley, 1999; Sweller, 1999).
- Active processing: Meaningful learning occurs when people engage in appropriate cognitive processing during learning, including attending to relevant incoming material, mentally organizing the material into a coherent cognitive structure, and integrating the material with relevant existing knowledge (Mayer, 2001; Wittrock, 1989).
Information Processing Model of Learning
Figure 18.1 presents a framework for describing the human information processing system based on the principles described above (Mayer, 2001). Information from the outside world—such as a textbook lesson or a teacher-led classroom demonstration—enters the learner’s cognitive system through the eyes and ears and is represented briefly in sensory memory. If the learner pays attention, some of the material is transferred to working memory for further processing (as indicated by the selecting words and selecting images arrows in Figure 18.1). Next, the learner may engage in deeper cognitive processing of the material in working memory, such as mentally organizing the material (as indicated by the organizing words and organizing images arrows) and integrating it with relevant prior knowledge from long-term memory (as indicated by the integrating arrow), and the resulting learning outcome can be stored in long-term memory.
Three Kinds of Memory Stores
As you can see, this information processing framework has three main memory stores: sensory memory, working memory, and long-term memory. Sensory memory is an unlimited but temporary store for holding incoming sensory information in which visual images and sounds last for a fraction of a second. Sensory information that impinges on the eyes is temporarily held as a fleeting visual image in visual sensory memory, and sensory information that impinges on the ears is temporarily held as a fleeting sound in auditory sensory memory. Working memory is a limited-capacity store in which a few pieces of incoming and retrieved material can be held and processed. Aspects of the visual images that are attended to are held in working memory as pictorial images and, when mentally organized by the learner, can be converted into a coherent pictorial representation (i.e., a pictorial model). Aspects of the auditory sounds that are attended to are held in working memory as sounds and, when mentally organized by the learner, can be converted into a coherent verbal representation (i.e., a verbal model). Further processing occurs in working memory as connections are built between the verbal and visual models with relevant knowledge retrieved from long-term memory. Thus, working memory is the venue for knowledge construction, but the amount of knowledge that can be held and the amount of processing that can take place at any one time is subject to capacity limitations. Long-term memory has unlimited capacity and is the storehouse for knowledge that has been constructed in working memory.
Three Kinds of Cognitive Processes During Learning
As you also see in Figure 18.1, this framework has three main kinds of cognitive processes: selecting, organizing, and integrating. By attending to aspects of material in sensory memory, the learner can transfer it to working memory for further processing. In Figure 18.1, selecting is indicated by the arrows from sensory memory to working memory. By mentally organizing the material in working memory, the learner can construct a coherent cognitive structure. Organizing is indicated by the arrows within working memory. By retrieving relevant prior knowledge and connecting it logically with incoming material in working memory, the learner can construct a meaningful learning outcome. Integrating is indicated by the arrow from long-term memory to working memory. In addition, the learner can make connections between corresponding aspects of the verbal and pictorial models as indicated by the arrow between them.
Active cognitive processing for meaningful learning requires that the learner engage in all three kinds of cognitive processing during learning—selecting relevant information, organizing it into coherent cognitive representations, and integrating it with other representations and knowledge from long-term memory. Finally, the arrow from working memory to long-term memory signifies a fourth kind of cognitive process—encoding the newly constructed knowledge in long-term memory.
Three Kinds of Cognitive Load
In the information processing model in Figure 18.1, each of the two channels in working memory is limited in capacity. Sweller (1999, 2005) and Mayer (2001, 2005a; Mayer & Moreno, 2003) identified three different kinds of cognitive load in working memory: extraneous cognitive processing (or extraneous cognitive load), essential cognitive processing (or intrinsic cognitive load), and generative cognitive processing (or germane cognitive load). Extraneous cognitive processing does not support the instructional goal and is caused by ineffective presentation format. For example, if a text lesson contains a lot of extraneous information and pictures, learners may focus mainly on that information instead of the key information. Similarly, in a group learning situation students may spend their time discussing topics that have nothing to do with the instructional task. An important instructional goal is to minimize extraneous processing, such as by eliminating extraneous material from the lesson.
Essential cognitive processing is needed to mentally represent the presented material (i.e., the process of selecting in Figure 18.1) and is caused by the complexity of the material to be learned. For example, a topic such as how lightning storms develop is complex because there are many interacting elements such as the effects of differences in air temperature and differences in electrical charge. An important instructional goal is to manage essential processing, such as by providing pretraining in the names and characteristics of the key elements or presenting the material in segments.
Generative cognitive processing is deeper cognitive processing (i.e., the processes of organizing and integrating in Figure 18.1) and is primed by the learner’s motivation to understand the material. An important instructional goal is to foster generative processing, such as asking learners to explain a lesson to themselves.
Overall, learners have a limited capacity for processing information in working memory, so instruction should be designed to minimize extraneous cognitive processing, manage essential processing, and foster generative processing.
Five Kinds of Knowledge
According to the information processing view, learning involves knowledge construction. An important contribution of the information processing view is the analysis of several qualitatively different kinds of knowledge:
- Facts: descriptions of things or events, such as “Earth is the third planet from the sun.”
- Concepts: principles or models (such as having the concept of place value for written numbers) and categories or schémas (such as knowing what a dog is)
- Procedures: step-by-step processes, such as knowing how to carry out the long-division procedure for 425 divided by 17
- Strategies: general methods, such as knowing how to summarize a paragraph
- Beliefs: thoughts about one’s cognitive processing, such as believing “I am good at learning about how the mind works.”
Learners possess all five kinds of knowledge in long-term memory, and proficiency in most complex tasks requires being able to coordinate among them (Anderson et al., 2001; Mayer, 2008). Metastrategies are strategies for how to manage and coordinate all the types of knowledge, and they are an important aspect of the knowledge of successful learners (McCormick, 2003).
Competing Views of Learning
Over the course of the past 100 years, researchers in psychology and education have posited four major metaphors of learning: response strengthening, information acquisition, knowledge construction, and social construction.
Learning as Response Strengthening
Learning as response strengthening—a popular view in the first half of the 20th century—conceptualizes learning as strengthening and weakening of associations. According to this view, the learner is a recipient of rewards and punishments and the teacher is a dispenser of rewards and punishments; a common instructional method is drill and practice. What is the relation of this view to the information processing view? In its traditional behaviorist form, the response strengthening view holds that rewards automatically strengthen associations and punishments automatically weaken them. In contrast, research in the information processing tradition has shown that it is not rewards and punishments per se that cause learning but rather it is the learner’s interpretation of the rewards and punishments (Lepper & Greene, 1978). In short, according to the information processing view, learners apply cognitive processing to the information in their environment including the rewards and punishments they receive.
Learning as Information Acquisition
Learning as information acquisition—a popular view in the mid-20th century—conceptualizes learning as adding information to long-term memory. According to this view, the learner is a recipient of information and the teacher is a dispenser of information; a common instructional method is lecturing or assigning readings. This is an early version of information processing theory in which information is seen as a commodity than can be transferred from one person’s memory to another person’s memory. In contrast, a more recent version—reflected in the knowledge construction metaphor—focuses on knowledge (which consists of cognitive representations in the learner’s memory system) rather than information (which consists of symbols that exist in objective reality for all to see) and focuses on the constructive processes (such as selecting, organizing, and integrating) rather than acquisition processes (such as simply adding information to memory). In short, many of the criticisms of the information processing view are attacks on the learning as information acquisition view, whereas the current version of the information processing view is reflected in the learning as knowledge construction metaphor.
Learning as Knowledge Construction
Learning as knowledge construction—a popular view since the last third of the 20th century—conceptualizes learning as building coherent cognitive structures. According to this view, the learner is an active sense maker and the teacher is a cognitive guide; an exemplary instructional method is asking learners to engage in self-explanation as they read or providing worked examples along with problems to solve. This view best epitomizes the information processing view of human cognition.
To better understand the distinction between learning as information acquisition and learning as knowledge construction, consider a lesson in which the teacher asks the class to view a 5-minute narrated video on how lightning storms develop. Which view of learning is this lesson most consistent with? It certainly is consistent with the information acquisition view because the instruction presents information for the learner to acquire. It would be consistent with the knowledge construction view only if the instructor also helps to guide the learner’s cognitive processing of the material—that is, if the instruction encourages the learner to select relevant information (such as focusing on key steps in the process), mentally organize it (such as building a causal chain of the key steps in the process), and relate it with prior knowledge (such as remembering knowledge related to each step, including why hot air rises). When the learners are inexperienced they may need some guidance in how to make sense of the material. This can be done by providing pretraining in the key terms (positively charged particle and negatively charged particle, freezing level, etc.), by highlighting key steps, by reminding students of their prior knowledge concerning temperature differences and differences in electrical charge, by breaking the lesson into segments that can be paced by the learner, by asking learners to explain each segment to themselves, and many other techniques that help learners process the material more deeply. Overall, the learning as information acquisition view is consistent with instruction that simply presents information to be learned, whereas the learning as knowledge construction view requires both presenting information and making sure the learner processes it appropriately.
Learning as Social Construction
Learning as social construction—an emerging view in the latter part of the 20th century—conceptualizes learning as a sociocultural event that occurs within groups as members work together to accomplish some authentic task. An exemplary instructional method is working together as a group on a significant academic project.
How does the learning as social construction metaphor compare with the information processing view? The answer depends on whether one takes a cognitive or radical approach. According to the cognitive version of social constructivism, people build cognitive representations when they work together on a task. This view is consistent with the basic tenets of the information processing view because individual learners are applying cognitive processes and building cognitive representations. According to a radical version of social constructivism, learning does not occur within learners and teaching does not emanate from teachers but rather is a sociocultural construction produced by a group and stored as a cultural product of the group (Phillips & Burbules, 2000). This view of social constructivism is not consistent with the information processing view, nor does it offer testable hypotheses needed to qualify as a scientific theory.
The information processing view has implications for learning, instructing, and assessing.
Implications for Learning
What is learning? Learning is a long-lasting change in the learner’s knowledge as a result of the learner’s experience. This definition has three components:
- Learning is long lasting, so fleeting changes such as mood changes do not count as learning.
- Learning is a change in knowledge, so changes in knowledge must be inferred from changes in behavior.
- Learning is a result of experience, so changes due to fatigue, injury, or drugs do not count as learning.
One of the most contentious aspects of this definition concerns the second element in the definition—namely, what is learned. According to the behaviorist view of learning, which dominated psychology and education in the first half of 20th century, what is learned is a change in behavior. According to the cognitive (or information processing) view of learning, which became dominant in the 1960s, what is learned is a change in knowledge. Knowledge is an internal, cognitive representation that is not directly observable, so it can only be inferred by changes in behavior. Thus, the information processing view focuses on changes in the learner’s behavior as a way of determining changes in the learner’s knowledge. A major implication of the information processing view is to modify the definition of learning so that what is learned is a change in knowledge rather than a change in behavior. The information processing view puts the construction of knowledge—or cognitive representations—at the center of learning.
Implications for Instructing
What is instruction? According to the classic view, instruction is concerned with presenting material to learners. In contrast, according to the information processing view, instruction is activity by the teacher intended to guide the learner’s cognitive processing during learning. The information processing model summarized in Figure 18.1 contains three kinds of cognitive processes during learning: selecting relevant information for further processing, organizing the selected information into coherent representations, and integrating the information with appropriate knowledge from long-term memory. Meaningful learning—the construction of a meaningful learning outcome—requires that the learner engage in all three kinds of cognitive processes. Rote learning—the construction of rote learning outcomes—requires that the learner engage in selecting relevant knowledge but does not require the deeper processing of organizing and integrating. Finally, no learning occurs when the learner does not engage in any of the three kinds of cognitive processes.
A major challenge of instructional design is to present material and encourage appropriate cognitive processing in a way that does not overload the learner’s information processing system. As you can see, a major implication of the information processing view is that the goal of instruction is more than simply presenting material; in addition, instructors must also guide the way that learners process the material. The information processing view changes the focus of instruction from presenting information to guiding the learner’s processing of the presented information. The information processing view puts cognitive processing—such as selecting, organizing, and integrating—at the center of learning.
Implications for Assessing
What is assessment? According to the classic view, the goal of assessment is to measure performance, such as how many arithmetic problems a learner can solve in a given period of time. The classic view of assessment is concerned mainly with determining how much is learned. In contrast, the goal of assessment in the information processing view is to measure knowledge, including the degree to which the learning outcome is meaningful or rote. The information processing view is concerned mainly with what is learned, that is, “knowing what students know” (Pelligrino, Chudowsky, & Glaser, 2001).
The information processing view has useful implications for how to assess what is learned. The two most common ways of assessing learning outcomes are retention tests—such as asking the learner to recall or recognize what was presented—and transfer tests—such as asking the learner to use the material to solve a new problem. No learning is indicated by poor performance on both retention and transfer. Rote learning is indicated by good performance on retention and poor performance on transfer. Meaningful learning is indicated by good performance on both retention and transfer. Thus, according to the information processing view, the quality of learning outcomes can be inferred by examining the pattern of performance on a series of dependent measures, including retention and transfer tests. Other techniques for probing knowledge use coding systems based on interviews and observations. Techniques for assessing one of the five kinds of knowledge (i.e., facts, concepts, procedures, strategies, or beliefs) may not be appropriate for assessing other kinds. Thus, a major implication of the information processing view is to assess what is learned rather than how much is learned.
Major Contributions in Educational Psychology
One way to judge the value of the information processing view is to examine whether it has generated useful research. In this section, I describe three examples of how the information processing approach has provided useful contributions to research in educational psychology: psychologies of subject matter, cognitive process instruction, and multimedia instructional design.
Psychologies of Subject Matter
Psychologies of subject matter represent a shift from studying learning in general to studying how learning works within specific subject areas. For example, instead of asking, How do people learn? researchers studying psychologies of subject matter ask, How do people learn to read, to write, to comprehend text, to solve math problems, or to think scientifically? (Mayer, 2004).
In order to help people learn to carry out basic academic tasks—such as reading a sentence, comprehending a paragraph, writing an essay, solving a word problem, or conducting a scientific experiment—the first step is to specify the cognitive processes involved in the tasks. Based on an information processing view, progress can be made by asking, What are the cognitive processes required for an academic task? and What do you need to know in order to accomplish an academic task? For example, in order to solve a math story problem, students need to be able to engage in four cognitive processes:
- Problem translation: converting each sentence of the problem into an internal mental representation by using factual knowledge (e.g., knowing how many cents are in a dollar) and linguistic knowledge (e.g., knowing that adding “s” turns a word into a plural)
- Problem integration: organizing the information into a coherent statement of the problem—called a situation model—by using conceptual knowledge (e.g., knowing place value) and schematic knowledge (e.g., knowing problem types)
- Solution planning and monitoring: devising a plan for solving the problem by using strategic knowledge (e.g., knowing how to break a solution plan into parts)
- Solution execution: carrying out the plan to arrive at an answer by using procedural knowledge (e.g., knowing how to add, subtract, multiply, and divide numbers)
Research on expertise (Bransford et al., 1999; Mayer, 2008) shows that experts—or proficient performers— know something different than novices. For example, successful mathematical problem solvers are able to represent the problem using concrete objects, and when less successful students are given instruction in how to represent problems in this way their problem-solving performance improves (Lewis, 1989; Low & Over, 1989). An important contribution of the information processing approach is the pinpointing of specific knowledge needed for proficiency on academic tasks (Kilpatrick, Swafford, & Findell, 2001). Overall, psychologies of subject matter represent one of educational psychology’s success stories in the late 20th century.
Cognitive Process Instruction
Cognitive process instruction involves providing focused instruction on cognitive processes needed for success on academic tasks. For example, if the goal is to improve reading comprehension, then students may need explicit instruction in how to engage in a process of self-explanation in which they try to explain discrepancies they find in the text as they read. Students who are taught how to engage in self-explanations show large improvements in their reading comprehension (Roy & Chi, 2005). Pressley and Woloshyn (1995) and Pressley and Harris (2006) have shown how cognitive process instruction can be applied across the curriculum. Thus, an important contribution of the information processing approach is the focus on teaching specific cognitive processes required for success in school. Cognitive process instruction has been another one of educational psychology’s success stories in the late 20th century.
Multimedia Instructional Design
Advances in instructional design principles represent a third example of the contributions of the information processing view. In particular, the information processing view has contributed to the creation of a new generation of principles for how to design instructional messages (such as textbook and online lessons). In taking an information processing approach, the focus is on helping learners engage in appropriate cognitive processing of the presented material, while being sensitive to characteristics of the human information processing system.
For example, some basic principles for how to design multimedia instructional messages (i.e., lessons containing words and pictures) are coherence, spatial contiguity, modality, and redundancy (Mayer, 2001; Mayer & Moreno, 2003). People learn better when extraneous material is excluded from a lesson. The coherence principle is based on the idea that working memory capacity is limited, so when a learner is processing extraneous material, the learner may not be able to process relevant material.
People also learn better when corresponding printed words and pictures are placed near rather than far from each other on the page or screen. The spatial contiguity principle is based on the idea that working memory is limited, so when the learner uses processing capacity to scan between words at the bottom of the screen and the appropriate portion of the graphic, the learner may have inadequate remaining capacity for deep processing of the target material.
The modality principal holds that people learn better from graphics and concurrent narration than from graphics and concurrent on-screen text. When learners must attend to both graphics and printed words they must split their visual attention (in visual sensory memory) because the eyes can only look at one location at a time. When words are presented in spoken form, they enter the information processing system through the ears (and auditory sensory memory), thus off-loading some of the processing from the visual channel.
Another principal of multimedia instructional design is the redundancy principle. People learn better from graphics and concurrent narration than from graphics, concurrent narration, and concurrent on-screen text. Adding on-screen text creates the same split-attention problem described for the modality principle and invites extraneous processing in which learners waste limited processing capacity on trying to reconcile the two verbal streams.
As you can see, these principles and many others (Mayer, 2005b) are based on an information processing view. The information processing view has contributed to yet another of educational psychology’s success stories—the development of theory-based instructional design principles that work when tested empirically (O’Neil, 2005).
The information processing view of learning holds that people construct cognitive representations by applying cognitive processes. The learner’s construction of knowledge corresponds to the “information” side of information processing, whereas the learner’s application of cognitive processes corresponds to the “processing” side of information processing. According to the information processing view, the main goal of education is to foster changes in the learner’s knowledge. This goal is accomplished by devising instruction that helps guide the learner’s cognitive processing during learning.
The information processing view has fundamental implications for 21st-century education. In the 21st century, the task of educators is not just to present information or to provide learning environments for students. In addition, a major challenge of educators is to guide how students process information during learning. To meet this challenge, educators would benefit from an understanding of how the human information processing system works.