Alan Baddeley's 1999 book describes the case of Clive Wearing, a talented musician who in 1985 fell ill with a viral infection that resulted in encephalitis which caused extensive brain damage. Wearing could no longer remember what happened more than a few minutes before. He continued to believe that he had just regained consciousness and kept a diary in which he constantly recorded that belief. Wearing could, however, still play music. Thus, some kind of memory remained intact. His case demonstrates the importance of memory.
In 1968 Atkinson and Shiffrin developed a model consisting of three different kinds of memory: sensory, short term, and long term (Atkinson & Shiffrin, 1968). Sensory memory is very brief, lasting about 1/3 of a second (Sperling, 1960). Short-term memory is temporary memory storage. It has limited capacity and duration. Long-term memory is a more enduring memory.
Miller described the capacity of short-term memory (STM) storage as 7 ± 2 bits of information (Miller, 1956). Contents in short-term memory last about 20 to 30 seconds unless an individual rehearses or elaborates on the material. There are two kinds of rehearsal: maintenance rehearsal and elaborative rehearsal. Maintenance rehearsal involves the repetition of the contents of STM; elaborative rehearsal involves a deeper form of rehearsal by which people connect the to-be-remembered information to what they already know. One way to reduce the demands on working memory involves chunking information, or grouping bits of information. For example, when trying to remember the phone number 555-1212, it easier to remember two chunks (“555” and “1212”) rather than the seven individual numbers. Information is lost from short-term memory through decay and interference. Decay occurs when information is not used, and it simply fades from memory. Interference occurs when something else gets in the way of recall.
The probability of recalling information presented in a list is influenced by the position of the information in the list, a phenomenon known as the serial position effect. Recency effects indicate that items from the end of a list will be recalled, and primacy effects indicate that words at the beginning of a list will be recalled. Recognizing the limits of short-term memory, teachers can keep their instructions brief and provide opportunities for rehearsal.
Working memory is a term used to describe a limited, though active, memory system. It differs from short-term memory in that it includes manipulation functions as well as storage. Baddeley proposed that working memory has a number of subsystems that are coordinated by a central executive (Baddeley, 1990). The first of these subsystems is the phonological loop system, which processes speech or auditory information. It consists of a passive phonological store and an articulatory rehearsal process.
The Phonological Loop System. The phonological loop processes verbal information. Evidence for it includes the phonological similarity effect, the unattended speech effect and the word length effect. People make more errors if the words they are asked to recall are similar in sound to one another (Baddeley, 1993). When individuals are asked to perform a verbal task (e.g., reading) with speech in the background, performance is impaired, a phenomenon known as the unattended speech effect. Verbal information that is presented auditorially is processed automatically. A third source of evidence in support of the phonological loop is found in the word length effect. There is a link between memory span and the length of words to be recalled.
The Visuospatial Sketchpad. A second subsystem in Baddeley's model is the visuospatial sketchpad, which is used for processing visual or spatial material or both. The results of a 1968 study by Brooks showed that when an individual engages in a visuospatial task such as pointing while performing a visual imagery task, the same processing capacity is being used. If the form of the task (verbal or visual) and that of the response (verbal or visual) are the same, performance is impaired.
The Executive System. The central executive system of working memory controls the phonological loop and the visuospatial sketchpad. It is an attentional control system with limited capacity (Baddeley, 1993). Daneman and Carpenter's 1980 study explored the relationship between working memory and reading comprehension by asking individuals to read passages that contained inconsistencies due to the presence of words with more than one meaning. Individuals with low working memory spans were able to come to the correct conclusion only 25% of the time. Individuals with high working memory spans were able to keep the initial information in working memory until they encountered the information that clarifies the passage. Additional research suggests that the functioning of the central executive system is the key difference between good and poor comprehenders (Oakhill, 1982, 1984; Oakhill, Yuill, & Parkin, 1986).
When instructions are complicated or lengthy, there is a risk that students will not remember them. If students fail to follow instructions, it may be that the instructions exceeded their working memory's capacity for processing information. Interruptions are frequent in elementary school classes and it seems reasonable to assume that the constant interruption can produce interference effects. Working memory is involved in such tasks as reading comprehension, writing, problem solving, and mathematics (Swanson & Siegel, 2001).
Individuals with a large working memory span utilize cognitive resources more efficiently while reading, and as a result have more resources for storage while comprehending the text (Swanson & Siegel, 2001). Students must also retrieve information from long-term memory to include in their writing. Maintaining ideas and choosing among them while actually producing text can make heavy demands on memory capacity. Some learners have difficulty writing because of limited working memory capacity.
Students with learning disabilities frequently have deficits in working memory (Swanson & Siegel, 2001). In particular, they have difficulty with reading comprehension because of deficits in the phonological loop. Such difficulties are problematic on tasks that require a learner to retain information in mind for a short period while also carrying out further activities. This skill is very important in reading tasks in which information that is coming in must be stored temporarily while other information is being processed (Swanson & Alexander, 1997). Difficulties with working memory can also interfere with writing, because efforts to record ideas may interfere with maintenance rehearsal in working memory.
Students and teachers can use a variety of strategies to reduce the demands on working memory. Children often count on their fingers, thus giving themselves a visible record of their cognitive activity rather than relying on memory. Other strategies for supporting working memory include presenting information in multiple modalities, allowing students to record their ideas before writing, or using speech-to-text software to reduce the burden on working memory. Using an external representation can reduce the demands on working memory.
Researchers have described several types of long-term memory. The Atkinson and Shiffrin model was based on the duration of memory. Processing was divided among sensory, short-term, and long-term memory. This model was very useful, but it does not provide a complete description of how memory works.
There are several ways to distinguish between various kinds of long-term memories. One important distinction is between episodic and semantic memory (Tulving, 1972). Episodic memory is memory of events and typically includes sensory information (things seen, heard, or smelled, and so on). Such memories often have heightened emotional content (happy, sad, fearful). These memories are embedded in a specific context—a specific time and place. In contrast, semantic memory is memory of verbal information or declarative knowledge—that is, knowledge about facts. For example, knowing one's address is an example of declarative knowledge. It is separate from sensory information and not tied to particular experiences. Combining episodic and semantic memories makes information more memorable and retrievable.
Semantic memory is memory for meaning and is thought to be organized like a network. In 1969 Collins and Quillian proposed the earliest network model of semantic memory. In this model, semantic networks are made up of a network of related propositions, the smallest units of meaning that can be verified as true or false. A proposition involves linking two concepts by a relationship. Connections between ideas also vary in strength and frequency of use. These factors are more important than the actual categorical structures.
A second distinction in types of memory is between declarative and procedural memory. Declarative memory is like semantic memory. For example, a person might have declarative memories about the structure of a bicycle. In contrast, procedural memory is memory about how to do something. In the example given above, Clive Wearing had retained procedural memory of how to play music.
Responding to a question about an image takes about the same amount of time as responding to a picture (Kosslyn, 1976). In one study, participants were asked to form a mental image of a cat. They were then asked questions such as, “Does the cat have a head?” and “Does the cat have claws?” Responses to the latter question took longer, as participants appeared to scan the image.
Image information is thought to be stored in piecemeal fashion in long-term memory. Images are created by activating the overall or global shape of the image; elaborations are then added to create a complete image (Kosslyn, 1980, 1983). More detailed images take longer to retrieve. Images are thought to be stored in a nonimage format that specifies a recipe for constructing the image. As with the processing of language, processing images takes time.
Semantic memory is organized in complex networks. Complex understanding of a domain will result in a dense network of interconnected propositions about that domain. Schemas refer to organized sets of propositions about a topic (Bartlett, 1932). Learners' available schemas influence how they interact with the environment. The learner's schemas may be altered as a result of interacting with the environment.
A schema that describes the typical sequence of events in a situation is called a script, or event schema. For example, when people go to a restaurant, they usually expect actions to unfold in a particular way. Another type of script is a story grammar that can help students to understand and remember stories (Gagné, Yekovich, & Yekovich, 1993).
In 1972 Craik and Lockhart proposed an alternative theory of memory. It suggested that memory differences are not so much a function of duration as of depth of processing (Craik & Lockhart, 1972). They showed that students who attended to meaning performed significantly better than those who attended to surface features of the words. Craik and Lockhart argued that the differences in performance reflected differences in depth of processing.
Craik and Lockhart's concept of levels of processing helped shift the emphasis in the study of memory from storage to processing. The Atkinson and Shiffrin model defined memory systems in terms of the storage/duration of memories and described very short stores (sensory and short-term memory) and very long-term stores (long-term memory). The levels of processing theory focuses on the likelihood of retrieval as a function of how effortful and meaningful the initial encoding was.
Encoding is the taking in of information. The probability that information will be retrieved or remembered depends on the quality of encoding. Key processes in good encoding are organization, practice, and elaboration. It is easier to learn organized material than it is to learn disorganized material.
Practice helps develop good memory. Material that is used more often is remembered more easily. However, there are different ways of practicing. Distributed practice is much more effective than massed practice. Distributed practice is done over a period of time, with varying intervals between rehearsals of the information. Remembering involves using the cues available to assist remembering but also involves generating cues that help remembering. Distributed practice allows students to practice both of these skills. Massed practice, in contrast, involves engaging in extensive practice at one time, such as studying all night before an exam. This can be somewhat effective for an immediate task but is unlikely to lead to long-term recall of information.
Elaboration also helps in encoding and retrieval. It involves connecting new information with prior knowledge or to images or other enhancements of the information to be learned. Images in particular are powerful aids to memory and are frequently used to elaborate on information.
Dual-coding theory (Paivio, 1986) explains why images are helpful in remembering. According to this theory, images and words are represented differently, as imagens and logogens. When the two forms of representations are linked, the memory for the information is stronger. Baddeley's findings on the separate working memory systems for visual and verbal information also support the importance of presenting and learning information in both visual and verbal forms (Baddeley, 1999). Visual strategies such as concept maps and graphic organizers integrate verbal, visual, and spatial information to enhance encoding and retrieval.
Remembering can occur through either recognition memory or recall. Recognition memory responds to cues. Responding to multiple-choice tests can require recognition memory because the cues provided by the options from which one may choose will provide some assistance to memory. When information is recalled, one must generate information without cues. Responding to an essay question, for example, requires that one generate and organize the content.
Retrieval occurs within a semantic network because activation is spread from one node to another. Related ideas are triggered. The specific linkages will vary from one person to another. Because of the structure of a network, retrieval can involve reconstructing the links between propositions and ideas.
Forgetting occurs when there is interference or decay. When material is encoded in an organized manner, more cues are encoded, thus making retrieval easier. Also, information that is used more frequently is easier to recall. Material that is not encoded in an organized way and not used often is more likely to be forgotten. This is so because in a propositional network model of memory, retrieval occurs through a process of spreading activation. When one node in a network is triggered, related nodes are also triggered as activation spreads along the links to them. When these links are used often, less activation is needed to generate the connecting node. If the nodes are highly interconnected with many links, forgetting is less likely, but retrieval may take longer. In a curious irony, the more people know, the longer it may take them to verify that they know it.
Forgetting occurs when information is not used. If students have not practiced solving geometry problems for a year, they are likely to have forgotten the steps in doing so. Forgetting also occurs because of interference.
A number of structures in the medial temporal lobes of the brain are important for memory. They include the amygdala, the hippocampus, and the rhinal cortex that underlies the amygdala and hippocampus. The hippocampus plays a key role in the storage of new memories (Gazzaniga & Heatherton, 2003). The hippocampus and surrounding rhinal cortex are the most important areas for the consolidation of memory (Eichenbaum, 2002; Gazzaniga & Heatherton, 2003). The frontal lobes are also considered important for memory; although people who experience damage to the frontal lobes do not suffer dramatic memory loss, they may have difficulty remembering the order of events. Brain-imaging studies show that when people try to remember a list of words, the frontal lobes light up (Buckner, Kelly, & Petersen, 1999). The frontal lobes are more active when a task requires deeper encoding.
Atkinson, R. C., & Shiffrin, R. M. (1968). Human memory: A proposed system and its control processes. In K. W. Spence & J. T. Spence (Eds.), The psychology of learning and motivation (Vol. 2, pp. 89–195). Orlando, FL: Academic Press.
Baddeley, A. (1990). Human memory. Needham Heights, MA: Allyn & Bacon.
Baddeley, A. (1993). Using your memory: A user's guide. London: Penguin Books.
Baddeley, A. (1999). Essentials of human memory. Hove, United Kingdom: Psychology Press.
Buckner, R. L., Kelly, W. M., & Petersen, S. E. (1999). Frontal cortex contributes to human memory formation. Nature Neuroscience, 2, 311–314.
Craik, F. I. M., & Lockhart, R. S. (1972). Levels of processing: A framework for memory research. Journal of Verbal Learning and Verbal Behavior, 11, 671–684.
Daneman, M., & Carpenter, P. A. (1980). Individual differences in working memory and reading. Journal of Verbal Learning and Verbal Behavior, 19, 450–466.
Eichenbaum, H. (2002). The cognitive neuroscience of memory. Boston: Oxford University Press.
Gagné, E. D., Yekovich, C. W., & Yekovich, F. R. (1993). The cognitive psychology of school learning (2nd ed.). Boston: Little, Brown.
Gazzaniga, M. S., & Heatherton, T. F. (2003). Psychological science. New York: Norton.
Kosslyn, S. M. (1976). Can imagery be distinguished from other forms of internal representation? Evidence from studies of information retrieval times. Memory and Cognition, 4, 291–297.
Kosslyn, S. M. (1980). Image and mind. Cambridge, MA: Harvard University Press.
Kosslyn, S. M. (1983). Ghosts in the mind's machine. New York: Norton.
Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 6, 81–97.
Oakhill, J. V. (1982). Constructive processes in skilled and less skilled comprehender's memory for sentences. British Journal of Psychology, 73, 13–20.
Oakhill, J. V. (1984). Inferential and memory skills in children's comprehension of stories. British Journal of Educational Psychology, 54, 31–39.
Oakhill, J. V., Yuill, N., & Parkin, A. (1986). On the nature of the differences between skilled and less skilled comprehenders. Journal of Research in Reading, 9, 80–91.
Paivio, A. (1986). Mental representation: A dual coding approach. New York: Oxford University Press.
Swanson, H. L., & Alexander, J. E. (1997). Cognitive processes as predictors of word recognition and reading comprehension in learning-disabled and skilled readers: Revisiting the specificity hypothesis. Journal of Educational Psychology, 89, 128–158.
Swanson, H. L., & Siegel, L. (2001). Learning disabilities as working memory deficit. Issues in Education, 7, 1–48.
Tulving, E. (1972). Episodic and semantic memory. In E. Tulving & W. Donaldson (Eds.), Organization of memory (pp. 381–403). New York: Academic Press.
- Coats and Car Seats: A Lethal Combination?
- Kindergarten Sight Words List
- Child Development Theories
- Signs Your Child Might Have Asperger's Syndrome
- 10 Fun Activities for Children with Autism
- Why is Play Important? Social and Emotional Development, Physical Development, Creative Development
- First Grade Sight Words List
- Social Cognitive Theory
- The Homework Debate
- GED Math Practice Test 1