Implications of Cognitive Load Theory for Multimedia Learning

Mayer, R.E. (Ed.) (2005). Cambridge Handbook of Multimedia Learning. New York: Cambridge. [Chapter 2]


Cognitive load theory for multimedia learning has been one of the theories that combines human cognitive process and instruction design .
Long term memory (LTM):
Large, unlimited
Rote learning and meaningful learning result in change in ltm.
Schema construction, cognitive constructs that allow multiple elements of information to be categorised as a single elements. Schemas are acquired over a long time and stored in ltm. 
For multimedia presentations, knowledge is held in schematic form in ltm.
If a schema has been acquired, practices over long periods of time can make it automatic, without conscious control.


Working Memory (WM):
Miller (1956), claimed that the number of elements that can be held in wm in a given time is 7 plus and minus 2.
The limited capacity of wm is constrained when dealing with new information
Baddeley (1992)'s wm model: a central executive and two subsystems (visuo/spatial sketchpad for vision like pictures and phonological loop for auditory material like speech) 
If these two subsystems are used for acquiring instructions, learning can be facilitated.


LTM and WM
Miller (1956)'s chunking: chunk together elements of information that could be processed in wm as a single element. Chunks reside in ltm or formed using information from ltm .
Atkinson ad Shiffrin (1968)'s model has sensory memory, wm and ltm, and information moves between the stores.
The way wm works is critically dependent on what information has been stored in ltm.
Understanding occurs when all relevant elements of information can be processed simultaneously in wm.: Elements are organized and combined into schemas held in ltm. When schema construction is progressed and the elements for understanding the topic is processed in wm, understanding occurs.


Schemas act as a central executive for wm. If they are not available, random generation followed by effectiveness testing occurs. 
If no schema, schemas by someone else can also be helpful to organize the information. So it would be the instructional guidance to help the learner construct schemas if he does not already have any. So the learner won't need a  random generation plus effectiveness test, which is time consuming.


Three types of cognitive load:
Extraneous: caused by inappropriate instruction design 
-worked example effect: reduces extraneous cognitive load by reducing the search 
-split attention effect: when attention must be split between multiple sources of visual information. Should integrate the sources physically in order to reduce the extraneous cognitive load  
-modality effect: when multiple sources are necessary for understanding. Extraneous cognitive load is reduced by presenting verbal material in spoken way instead of printed text--using dual modality, which increases the effectiveness of wm
-redundancy effect: some sources of information can be redundant and unnecessary. Should be eliminated in order to reduce extraneous cognitive load.
-expertise reversal effect: information that is essential for novices becomes redundant for expert learners.


Intrinsic: natural complexity of the information. 
Germane: effective cognitive load, caused by effortful learning resulting in schema construction and automation. 


The aim of the instructional design should be to reduce extraneous load in a complex information which already has high intrinsic load.