V-+Information+Processing

= Cognitive Information Processing =

toc
 * The Cognitive Information Processing (CIP) model** of learning offers insights into how individuals acquire and incorporate knowledge. As the name suggests, the basis of the model focuses on the processing aspects of the mind. At its most basic level, learning occurs when input information from the environment is transferred to memory (Driscoll, 2000, p. 76). The model theorizes that information is processed in a similar way as a computer would, in discrete stages--from sensory memory to working memory to long-term memory. The model also attempts to describe each of these processing areas.

Unlike the behaviorist model, the CIP model emphasizes understanding the internal processes associated with learning. This model defines sensory information, as well as short-term and long-term memory. This model also includes possible explanations of how memory is stored. To begin the overview, I'll first begin by describing the stage model of memory.

The Stage Model of Memory


Based on the work of Atkinson and Shiffrin (1968), this model of memory is the core of the CIP model. After sensory stimuli are processed, the CIP model suggests that learning is facilitated by the interplay between working memory, or short-term memory (STM), and long-term memory (LTM). Although the two work together, there are key property differences between the two in the areas of capacity, code, permanence, source and loss (Driscoll, 2000, p. 78).

Capacity
From a learning standpoint, capacity refers to the amount of information that can be held in an individual’s memory at any given moment (Ibid., p. 88). In general, STM has a smaller capacity than LTM. George Miller (1956) showed that STM is limited to 7 +/- 2 numbers. More recent studies have show that the capacity of STM can be effectively increased through the process of chunking (Driscoll, 2000, p. 89-90). Unlike STM, LTM has a large capacity and is not limited in the same way as STM. The amount of information an individual can hold in LTM is not limited by capacity, but rather, by other factors such as interference or retrieval and encoding failures (Ibid., p. 104-105). Those other factors lead to the next area of difference between STM and LTM: encoding.

Code
Encoding describes the process of how individuals assimilate new information. The CIP model suggests the processes of information encoding in STM and LTM are not the same. For STM, the model indicates that information is encoded in a dual manner. As individuals process information in STM, they combine the new information with existing information (Ibid., p. 91). Unlike STM, LTM encodes information as either episodic or semantic. The two encoding mechanisms of LTM refer to information that is either based on an event or classified as "general" information (Ibid., p. 91-92).

Permanence
The third difference between the two memory processing units has to do with the concept of permanence. Permanence refers to the length of time the information is held by the processing unit. An individual continually receives information as he interacts with the world. In the CIP model, information is held in STM for only 20-30 seconds before it is lost, thus the information that is held in STM is constantly being supplanted by new information (Ibid., p. 90). In regards to LTM, no such limitation exists. Information maintained in LTM is permanent and is constrained by the same factors that limit capacity.

Source
Another difference between STM and LTM is the source of the information that is to be processed. Unlike STM, where the source of information is from the environment or prior knowledge, the source of information in LTM is from STM that has been encoded (Ibid., p. 78). As information is processed by STM, the CIP model theorizes that the output of this information is stored into LTM. A variety of ideas have been proposed to more deeply understand how the process of storage in LTM occurs, but all are based on the notion that the source is relayed from STM (Ibid., p. 94).

Loss and Retrieval
Loss refers to the way in which information stored in memory disappears. As described in the section on permanence, information residing in STM has a short shelf life. The information stored there either gets displaced by new information or simply fades away (Ibid., p. 79). By contrast, for LTM, information loss occurs due to a recall failure. The same factors that limit permanence and capacity in LTM, also affect loss in LTM. The opposite of loss is retrieval. Retrieval of information occurs at a rapid pace in STM because of issues relating to capacity, permanence and source. STM retrieves information from LTM, but only if that information has been effectively encoded. The concept of retrieval is highly relevant to LTM as it is where the CIP model believes much of the information an individual knows is stored. Retrieval of information can be split into two categories: recall and recognition. The difference between the two is in the use of cues. Recall refers to information that is retrieved without cues, while recognition does (Ibid., p. 101).

Models of Memory Storage
The CIP model also attempts to investigate how semantic memories are retained in long-term memory (Driscoll, 2000, p. 94). A variety of models have been proposed such as, the Network Model, the Feature Comparison Model, the Propositional Model, the Parallel Distributed Processing Model, and the Dual-Code model. Each of the models attempts to theorize how knowledge is stored.

The Network Model is based on the concept of memory nodes and describes memory as the arrangement of mental concepts (Ibid., p. 94). The Feature Comparison model is also based on the memory nodes, but instead of specific concepts, the nodes represent generalized traits (Ibid., p. 96). The Propositional Model is similar to the Feature Comparison Model but goes a step further and speculates that memories are encoded as in whole propositions (Ibid., p. 96). The Parallel Distributed Processing model explains memories as the relative strength of connections between a wide array of memory nodes (Ibid., p. 98). The Dual-Code Model believes verbal memories are encoded with non-verbal memories (Ibid., p. 99).

The CIP Model in Action
media type="custom" key="2554223"

This short video from [|TeacherTube] is based on Mike Rutherford’s [|Learning Centered Schools] professional development program for teachers. The video is an example of how a teacher might gain the learner’s attention based on principals of the CIP model.

Analysis and Discussion of the CIP Model's Application to Instruction
-

Like all learning theories, the CIP model cannot perfectly describe the entire learning process. However, the model provides concepts that can be directly applied to a variety of instructional settings and serves as a foundation for further research. Eric Gagne’s theory of instruction is based on the CIP model. In his model, there are three main components: a taxonomy of learning outcomes; conditions of learning, and the nine events of instruction (Driscoll, 2000, p. 341). Gagne’s integrated model of instruction serves as a comprehensive application of the CIP model.

General Principals
A more generalized interpretation of the CIP model’s instructional implications are in these three areas: providing organized instruction, arranging expensive and variable practice, and enhancing learners self control of information processing (Ibid., p. 106).

Organized Instruction
The CIP model believes that for instruction to be successful, it must consider the learner's attention. Since the learner's attention capacity is limited, the instruction must be organized in a way to maximize relevant information. This concept impacts instruction as it is being delivered. If the learner does not know what to pay attention to, then the learner might be overwhelmed with information.

To help focus the learner’s attention, the instructional tactic of signaling might be used (Ibid., p. 106). A graphical signal organizes instruction to aid learners in the identification of appropriate patterns in the lesson. Once the learner identifies the pattern, he will have an easier time making sense of the lesson. These patterns, or schemes, helped the learner structure knowledge.

Extensive and Variable Practice
The CIP model, asserts that learning occurs by assimilating new concepts and building upon previously learned concepts. One way to activate the process of building upon previous learned concepts is through extensive and variable practice (Ibid, p. 109).

Through extensive practice, the learner is exposed to a multitude of practice problems. This practice helps the learner gain fluency with new concepts. Through variable practice, the learner is exposed to a wide variety of practice problems. By learning how new concepts can be applied to a variety of subject areas, the learner is able to gain an understanding of complex concepts.

Driscoll (2000) uses the example of learning scientific concepts such as fusion, cohesion and adhesion by showing how those concepts apply to subject matter classes like science, technology, and art. The variety of practice helps the learner gain a deeper level of understanding then if the learner were only exposed to one type of subject matter class.

Self-control of Information Processing
In the CIP model, information is constantly being processed from short-term memory to long-term memory. That interplay is controlled through an executive function. The concept of executive control is similar to attention, but has a more general application.

For instruction, executive control relates to learning strategies (Ibid., p. 110). By providing guidance to learners on how to acquire and organize information, the learner is better able to assimilate information. An internal framework for acquiring information is related to the concept of Metacognition, or thinking about thinking.

Some learning strategies are quite simple, such as breaking down complex problems into a set of smaller ones. Another simple learning strategy is note-taking during a lecture. A more complex learning strategy is brainstorming ideas prior to writing an essay. By teaching learning strategies, in addition to facts and concepts, the learner is more likely to retain the information being taught.

Instructional Shortcomings
Three shortcomings of the CIP model are its complexity, cutting-edge nature and generalizability of laboratory research. The CIP model provides insights into the mind, but might be difficult for an educator to translate into the instructional setting. Also, the knowledge gleaned from how the brain functions might conflict with current instructional techniques. As brain science develops, new discoveries might contradict previous research. The emerging nature of the field inherently means that paradigms are not set and likely will shift over time. Many of principals that researchers have uncovered about memory have arisen from laboratory studies involving word lists (Barsalou, 1992, p. 146). These principals might not be applicable to everyday knowledge.

An example of a conflict with accepted instructional practice is in the area of classroom visual aids. The trend toward a creating rich classroom environment might go against the CIP model’s belief in reducing distraction and limiting the external cognitive load while learning. Rich classroom environments provide learners with a variety of stimuli that has the potential to overwhelm their senses and reduce their ability to encode information.

Up-and-coming research on the brain is leading to new discoveries on a seemingly daily basis. One area that is currently gaining attention is in the dynamic systems aspects to cognition (Beers, 2000). In recent years, there has been a growing body of evidence to suggest that it is impossible to reduce learning to lab simulations. Rather, learning must be studied from a dynamic systems approach. This trend may cause previous cognitive models, such as the stage theory of memory, to become outdated. Since the stage theory lies at the heart of the CIP model, this is a clear limitation of its instructional implications.

Personal Understanding
One of the reasons I chose this topic was because of my undergraduate degree in cognitive science. After graduating, I worked as a programmer to pay the bills, but have always been fascinated by the field of learning and cognition. It was only when I entered the ITEC program that I’ve been able to put my background to use.

An aspect of the Information Processing theory that appeals to me is the quantifiable nature of it and its direct relevance to education. Unlike other theories that lack a biological basis, this one can be studied from a psychological, neuro-scientific, and computational vantage point. The fact that principals can be verified through a variety of methodologies strengthens their theoretical relevance to learning. The theory provides a lens for educators to view how their students learn and can help them design more effective instruction.

As I study the field of Instructional Design, I can already see how this theory shapes the way I approach the design process. From not only the human factors point of view, but also the overall architecting of information, this theory provides me with a structure for creating both computer-based and instructor led trainings. By factoring in the underlying principals of the theory, the result is a more soundly built project.