Samiksha Jaiswal (Editor)


Updated on
Share on FacebookTweet on TwitterShare on LinkedInShare on Reddit

The Limited Capacity Model of Motivated Mediated Message Processing or LC4MP is an explanatory theory that assumes humans have a limited capacity for cognitive processing of information, as it associates with mediated message variables; moreover, they (viewers) are actively engaged in processing mediated information Like many mass communication theories, LC4MP is an amalgam that finds its origins in psychology. Specifically, this theory has its origins in the Limited Capacity Model for understanding cognitive information processing. The most fundamental assumptions of information processing are the three dimensions of cognitive processing. The three dimensions: 1) encoding, 2) storage, and 3) retrieval. This is how viewers get presented information into their heads. Messages can be processed under controlled conditions or they can be automatically elicited.


Limited capacity theory

When a message processor decides to pay attention to a message because it appeals to their interests, and they allocate resources to information processing, the controlled message engagement subprocess begins; conversely, when a message processor is cued automatically, and they are paying attention, the same process of allocating cognitive resources begins to elicit message processing. First, message engagement, which is a stimulus approach/avoidance interaction engages the appetitive and aversive cognitive subprocesses. In the most lay terms, these are basic fight or flight responses that happen in mere nanoseconds. This information then can report to sensor stores in the brain; and if it is useful, it will move to short term memory and long term memory.

Encoding, storage, and retrieval

Encoding of information is a subprocess where a stimulus orients a person to a message, vel non. In terms of message processing, its effectiveness is determined by the number of cognitive resources allocated to the encoding process. If the person is exposed to a stimulus, and they choose to pay attention, the brain will treat that mediated message like it was reality. That is, the brain creates an "orienting response (OR), which is an automatic cognitive response to novel information". This response is a cognitive reaction to a change in the environment, and can be measured via implicit measures of recognition and explicit measures such as electrochemical reactions in the body. These cognitive processes can be both automatic, as in the case of OR’s, or controlled. In either event, the next step in the message engagement process is to move the information into storage.

Storage is conceptually modeled as a cognitive sub-processing of information into and through the general associative network. Associative networks are conceptual “nodes” that map brain uses as sensor stores—for short-term and long term memory. They are conceptualized as the site where information is stored and ultimately retrieved. Under normal conditions, if there are many associative links available for information processing, “the more readily retrievable that information will become”; hence, then “retrieval” time of information is faster and more accessible. Storage of information is measured via cued recall.

Information retrieval is conceptualized as the third fundamental conceptual sub process of information processing, and it is a function of memory, which is measured via free recall. Measuring free recall is based on an individual’s levels of sensitivity to stimuli, without being cued to the information.

Cognitive memory

Automatic cognitive responses are elicited via a cognitive response to novel information, the primary method in which automatic responses are elicited is through structural cues of the communication situation. Lang, et al. (2000) elaborates on this by controlling the structural features. They discovered that the relatedness of the media messages mattered. They dissected the pacing of mediated content into two categories: one that used related editing in camera angle and content; and, another with unrelated cuts to unrelated information to the previous information that was being encoded and stored. These pacing cues are important to media producers because they can either help reinforce attention into memory; or they can created a low memory retention condition in the minds of the message processor.

By inserting edits, which are operationalized by the process of switching cameras or transitions in visual or audio media, media producers can elicit orienting responses in attentive viewers. This is accomplished because the mechanism of allocating resources to memory message processing is effective in low, medium, and high edit conditions. Therefore, retrieval is also possible and is measured via recognition testing. However, unlike related editing cues, unrelated cuts do not make it into memory, especially when the cuts are coming at the message processor rapidly. This is important, secondarily, because the entire illusion of a commercially driven media system is premised on the concept that advertising sells products. More importantly, this article's primary concern is to instruct media producers—who are concerned with generating public information, the importance of editing cues, and the deleterious uses of disparate advertising and its ineffectiveness.


Methodologically, the LC4MP uses three primary sets of measures to study how people allocate resources and process information. Secondary task reaction time (STRT) is a measure of the amount of resources allocated to attention. STRT involves people engaging in a primary task such as watching TV while also monitoring a second task such as pressing a button whenever a cue such as a tone sounds or a light comes on. The amount of resources that are available (i.e., resources allocated minus resources required) has been demonstrated to influence how long it takes people to respond in the STRT. When there are a lot of resources available, people quickly notice the light or tone and are quick to press a button. But when resources are low or taxed, people take longer to notice the light or tone and are slower to respond. For example, a cut to an unrelated scene tends to tax resources because of the new information that needs to be processed, and STRT is slower. Critically, STRT is a function of both the resources allocated to the task and the resources required by the task.

A second set of methodological tools involves physiological measures such as heart rate, galvanic skin responses (sweating) or muscle activity. Heart rate is used as a measure of whether an orienting response has occurred. When people show an aroused response, their heart rate decelerates for a few seconds before returning toward their baseline. Likewise, an increase in skin conductance is evidence of an aroused response. Thus, the combined evidence of heart rate deceleration and an increase in skin conductance provides convergent evidence that an orienting response (OR) has occurred.

A final set of methodological tools involves measures of memory including recall and recognition measures. The measures of memory are used to make inferences about encoding and retrieval processes. For example, if messages are simple, increasing people’s attention to a message through the inclusion of formal features that orient attention, such as edits, results in improvements in memory for the message. This finding suggests that there were available resources (i.e., more resources allocated than required) that could be diverted to encoding processes. Conversely, with a complex message, the addition of features that increase ORs tends to lead to decrements in memory, which suggests that ORs automatically drew resources away from encoding processes and resulted in worse memory. The decreased performance on memory tests is one indication that the resources required has exceeded the resources allocated to the task.

Limits of LC4MP

The ability of information processing of mediated messages has limits. Interference with the encoding, storage, and retrieval of messages; especially when the receiver is not interested in the message, creates a situation where the required resources needed to process information are not dedicated; and, “fewer associative links will be made available to the message... therefore, it limits their ability to process media” (). This is a concern for researchers interested in studying how people encode and store information because if information processors do not devote the appropriate number of resources required to process information, then information overload will inhibit the abilities of individuals to encoding, store, and retrieve information. Additionally, the ability to information processors to process messages is also limited by their unique retrieval sub processing abilities, which are related to perceptions compiled over time. Thus, the sub process of retrieval is affected by two dimensions termed, “later retrieval-done after the fact and concurrent retrieval- in real time affect information processing of messages”.

As conceptual level of LC4MP, constructivism argues the limitation of information processing theoretical view, although the argues are not in mediated media world, but in more foundational psychology world. Before to point out specific limits of the view, it is useful to summarize the characteristics of information processing views for contrast. First, this view has it that information gains (e.g., learning) can be conceived as internal structures (e.g., long term memory) changing to reflect changes in the external world, which represents "reality". Within the traditional information processing view this real world is "knowable" and, as a consequence a student's knowledge can be ultimately judged as right or wrong. A second important characteristics of the information processing view is that there are general principles of learning, knowledge, etc. that hold true across situations and across learners, and it is the job of the researcher and theorist to ascertain these general principles.

On the other hand, a fundamental characteristic of the prototypical constructivist view is that no knowledge is absolute, rather knowledge is constructed. For example, the constructivists point to all the times in history when something was thought be a "fact", but was later discounted. This view is in contrast with the "absolute reality" basis of the information processing view. Another fundamental characteristics of the structuralist perspective is that learning only occurs within some context. So, for example, to consider learning independent of the environment in which it occurs (e.g., school, work, or play), is unrealistic and incomplete. This, again, is inconsistent with both the behaviorist and information processing views (and much of science in general, for that matter) in that these latter theories are based on the assumption that fundamental principles can be ascertained that will apply broadly across contexts and across learners.

These distinctions may sound quite abstract and removed from "real life" classroom instruction to you. In fact, in a practical/applied sense there are some important implications and practices that have emerged from this constructivist view since it has come into favor in the last decade. The general view that knowledge is not so absolute, that it is "constructed" puts more of an emphasis on the knowledge and conceptions that students brings to the classroom. This leads to a more "learner centered approach", with the teacher as more of a "guide on the side", as opposed to a "sage on the stage". Further, in a "constructivist classroom" students are more free to participate in the planning (i.e., "negotiate") and, even, the assessment process. Another implication of the "relative" nature of knowledge contention is that it is important to present the learner with multiple perspectives on a single topic or concept.


LC4MP Wikipedia

Similar Topics
The Desert Song (1929 film)
Kathrine Narducci
Brian Dickinson