|
|
|
|
|
| Contents |
|
Abstract
2 The theory of Sweller and Chandler 9 Hot links |
| Abstract |
|
Sweller and Chandler show how more efficient instructional designs can be gained from
the analysis of cognitive load (intrinsic and extraneous).
|
| 1 Literature |
|
Sweller and Chandler 1994
John Sweller and Paul Chandler: Why Some Material Is Difficult to Learn. In: Cognition and Instruction, 1994, 12 (3), 185-233. Lawrence Earlbaum Associates, Inc.
Sweller and Chandler 1991
|
| 2 Sweller and Chandler: The cognitive load theory |
| Assumptions and suggestions |
# Assumption 1
The major learning mechanisms when dealing with higher cognitive activities are:
The two mechanisms help to circumvent our limited working memory and emphasize our highly effective long-term memory. # Assumption 2 Beacuse our working memory is limited, we have a hard time to assimilate multiple elements of information simultaneously. # Assumption 3 But under conditions where multiple elements of information are interacting, we have to assimilate them simultaneously, in order to learn them together. # Assumption 4 Material with a high level of element interactivity produces a heavy cognitive load. # Assumption 5
Causes of high levels of element interactivity and their associated cognitive loads may be:
# Assumption 6
Cause of a heavy extraneous cognitive load may be:
The extraneous cognitive load does not disturb as long as the intrinsic element interactivity and consequent cognitive load are low - but is critical, as soon as we have to learn with intrinsically high element interactivity materials. Suggestions Sweller and Chandler suggest their way to classify our difficulties in assimilating information:
# natural origins (determined by the nature of the materials);
To deal with these difficulties they suggest specific procedures and support their instructional design with empirical data.
|
| Schemas and automation relieve or bypass working memory |
Some relations among long-term memory, working memory, schema acquisition, and automation
The main source of intellectual skill is the long-term memory. See the studies of expert-novice differences: De Groot's, 1965; Sweller & Cooper, 1985; Jeffries, Turner, Polson & Atwood, 1981; Egan & Schwartz, 1979. Our limited working memory allows to deal with no more than about 7 items of information at a time (Miller, 1956), or even less (Simon, 1974).
We reduce the burden of cognitive load in the working memory by:
A schema is a cognitive construct that organizes information
according to the manner in which it will be dealt.
As soon as we have built up a tree schema in our long-term memory, we do not have to store the immense detail of information presented by a tree. Such a load would overwhelm our working memory. Automation allows cognitive processes without conscious control. A schema will become automated after considerable practice. The learning mechanisms schema acquisition and automation use material stored in long-term memory to reduce the burden on working memory.
|
| High element interactivity leads to heavy cognitive load |
Element interactivity as an intrinsic source of cognitive load
Learning some vocabulary of a second language asks for a low cognitive load,
because many words can be learned in isolation from the other words (not always true, C.).
Learning difficulty is a function of the number of elements that must be learned simultaneously.
The degree of interactivity of the elements varies from one subject matter to the next.
Greater element interactivity means heavier cognitive load. See: Halford, Maybery & Bain, 1986; Maybery, Bain & Halford, 1986. Relations to schema theory
When dealing with information consisting of many interacting elements, we are dealing with schemas.
Measuring Element Interactivity
The expertise of the individual who is learning the material determines what are the constituent elements
(that must be considered simultaneously) in the task, and on which level of expertise they act.
|
| Instructional design can facilitate or impede learning |
Instructional format as an artificial source of cognitive load
The difficulty of an area is determined by:
The cognitive load imposed by the intrinsic nature of the material is determined by:
An extraneous cognitive load is one that is imposed purely because of the design and organization
of the learning materials rather than the intrinsic nature of the task - due to:
Effects that have given rise to techniques designed to reduce extraneous cognitive load (with empirical evidence of effectiveness):
The split-attention effect
If instructional material unnecessarily requires students to split their attention among multiple sources of information,
an extraneous cognitive load is imposed.
The redundancy effect
If instructional material unnecessarily brings elements together that can be understood in isolation without any problems,
element interactivity will be increased.
|
| Cognitive load theory |
Cognitive load consequences of relations between element interactivity and instructional design
The learning mechanisms (schema acquisition and automation) reduce the burden on working memory
by emphasizing long-term memory.
When dealing with material that has a high level of intrinsic element interactivity, presentation techniques become important: extraneous cognitive load (caused by split-attention effect or by redundancy effect) may be overwhelming. In contrast, as long as dealing with material that has little or no intrinsic element interactivity, presentation techniques may not matter as much - even heavy extraneous cognitive load can be treated in the working memory.
|
| Hypotheses |
Learning to use equipment
Sweller and Chandler organized a quite interesting setup to test their theory.
For this purpose they built up 3 groups: # Conventional-manual-plus-computer group: worked with the conventional software manual and the computer, followed the conventional manual and executed the activities (keyboard, screen). # Modified-manual-only group: worked with a modified manual that included diagrams of screen information and keyboards, followed the modified manual only, did not touch the computer. # Modified-manual-plus-computer group: worked with the same modified manual as the modified-manual-only group, followed the modified manual and executed the activities (keyboard, screen).
Sweller and Chandler expected the following results -
if the computer application that must be learned has a high degree of intrinsic element interactivity:
In both cases, whether the modified-manual-only group is superior should depend on the extent to which the information learners are attempting to assimilate has a high or low degree of intrinsic element interactivity. The effects should only be obtainable with high levels of intrinsic element interactivity.
|
| Experiment 1 |
20 persons had to learn several tasks of a CAD/CAM software package
(computer-aided design, computer-aided manufacture).
Two groups:
The conventional-manual-plus-computer group required learner
to split their attention among the manual, screen and keyboard.
Method The subjects were 20 first-year trade apprentices from a Sydney company: at least 10 years of high school, enrolled at technical colleges, previous experience with computers, no previous exposure to CAD/CAM programs.
The instructional materials consisted of two sets of manual instructions: conventional and modified.
The experiment was conducted in two phases:
Results and discussion
The variables under analysis were:
The tasks with low element interactivity showed some superiority of the modified-manual-only group in time and score - with no significance. But in the tasks with high element interactivity the modified-manual-only group worked much better than the conventional-manual-plus-computer group: in time and score - with strong significance. Even in practical test, where the conventional-manual-plus-computer group had considerable exposure to the computer during the instructional phase, the modified-manual-only group was much better - in time and score. Despite spending less time studying their instructions, the modified-manual-only group was superior to the conventional-manual-plus-computer group in both written and practical skills - as long as dealing with high element interactivity. Access to the computer interfered with learning (split-attention effect), while no exposure to the computer proved to be an advantage.
|
| Experiment 2 |
30 persons had to learn several tasks of a commonly used spreadsheet software package.
Three groups:
The conventional-manual-plus-computer group required learner
to split their attention among the manual, screen and keyboard.
Method The subjects were 30 high school students, 7th grade, from Sydney high school: previous experience with computers during primary school, no previous exposure to spreadsheet programs.
The instructional materials consisted of two sets of manual instructions: conventional and modified.
The experiment was conducted in two phases:
Results and discussion
The variables under analysis were:
The tasks with low element interactivity showed little difference among the three groups. But in the tasks with high element interactivity the modified-manual-only group outperformed the two other groups, which did not differ from each other - with strong significance. Even in practical test, where the two other groups had considerable exposure to the computer during the instructional phase, the modified-manual-only group worked much better . Despite spending less time studying their instructions and despite having had no previous exposure to the computer before testing, the modified-manual-only group outperformed the conventional-manual-plus-computer and the modified-manual-plus-computer group in both written and practical skills - as long as dealing with high element interactivity.
Again, access to the computer interfered with learning,
while no exposure to the computer proved to be an advantage.
A self-contained, modified manual designed to reduce extraneous cognitive load displayed its superiority over other presentation formats by eliminating split-attention and redundancy.
|
| Experiment 3 |
30 persons had to learn several tasks of a commonly used word-processing software package.
Three groups:
The conventional-manual-plus-computer group required learner
to split their attention among the manual, screen and keyboard.
Method The subjects were 30 high school students, 7th grade, from Sydney high school: previous experience with computers during primary school, no previous exposure to word-processing programs.
The instructional materials consisted of two sets of manual instructions: conventional and modified.
The experiment was conducted in two phases:
Results and discussion
The variables under analysis were:
There were only tasks with low element interactivity and they showed no significant difference among the three groups - in written and in practical test. The modified-manual-only group spent significantly less time working thru their instructions than the other two groups, since they did not work on the computer. Because most elements of the word-processing package could be learned in isolation, there was little element interactivity and light cognitive load. Under these circumstances, any extraneous cognitive load imposed by the presence of the computer was not an important factor.
|
| Experiment 4 |
20 persons had to learn several tasks with noncomputing technical equipment
(electrical engineering materials).
Three groups:
The conventional-manual-plus-apparatus group required learner
to split their attention among the manual and the technical equipment.
Method The subjects were 30 first-year trade apprentices from a Sydney company: at least 10 years of high school, enrolled at technical colleges, previous experience with the technical equipment, no previous experience with the tasks of the practical test (testing of electrical appliances). The instructional materials consisted of two sets of manual instructions: conventional and modified.
The experiment was conducted in two phases:
Results and discussion
The variables under analysis were:
The tasks with low element interactivity showed little difference among the three groups. But in the tasks with high element interactivity the modified-manual-only group outperformed the two other groups, which did not differ from each other - with strong significance. Even in practical test, where the two other groups had considerable exposure to the the technical equipment during the instructional phase, the modified-manual-only group worked much better . Despite spending less time studying their instructions and despite having had no previous exposure to the technical equipment before testing, the modified-manual-only group outperformed the conventional-manual-plus-apparatus and the modified-manual-plus-apparatus group in both written and practical skills - as long as dealing with high element interactivity. Their superiority was demonstrated in all areas of testing, including performance on the two transfer problems. On some tests, the difference among the groups were very massive: mean score from three to five times greater.
Like in the other experiments, access to the technical equipment interfered with learning,
while no exposure to the technical equipment proved to be an advantage.
Once again, a self-contained, modified manual designed to reduce extraneous cognitive load demonstrated its superiority over other presentation formats by eliminating split-attention and redundancy effects.
|
| General discussion |
Learning to use equipment might be facilitated by the absence of the equipment!
Information can be difficult to assimilate because:
If elements interact and cannot be considered in isolation, we are forced to process them simultaneously. What constitutes an element and which element must interact when learning a task are entirely dependent on the schemas that have been acquired by learners. An integrated schema can be used as an element in other, more advanced contexts. Levels of element interactivity are critically determined by expertise, which in turn is determined by the extent of schema acquisition.
If learners must simultaneously assimilate and mentally integrate elements
in a manual and on a computer screen or associated with physical apparatus,
cognitive load is likely to be higher than if all of the material is physically integrated in a manual.
Learning to use equipment from a manual alone can be easier and more effective than learning from a manual plus the relevant equipment.
Proceeding the other way round and presenting all instructional material on the computer screen
(like in CBT software)
rather than in a printed manual might show similar effects - not investigated here.
These experiments do not cover spatial-motor coordination: tasks that include significant spatial-motor components are probably learnable only by carrying out the physical activity.
|
| Links |
|
Michael J. Albers: Phd
Reading List
Laurie Cestnick: The Art of the Chart: improving discourse reading comprehension of adults with brain injuries. Paul Chandler: Learning with software: pedagogies and practices. Is conventional computer instruction ineffective for learning? Graham Cooper: Cognitive load theory as an aid for instructional design. Australian Journal of Educational Technology, 1990, 6(2), 108-113. Meredyth Daneman, Philip M. Merikle: Working Memory and Language Comprehension: A Meta-Analysis Psychonomic Bulletin & Review, 1996, 3, 422-433 Fischer, F., Kittel, A., Grδsel, C. & Mandl, H. (1996). Strategien zur Bearbeitung von Diagnoseproblemen in komplexen Lernumgebungen (Forschungsbericht Nr. 66). München: Ludwig-Maximilians-Universität, Institut fόr Pädagogische Psychologie und Empirische Pädagogik, Lehrstuhl fόr Empirische Pädagogik und Pädagogische Psychologie. Foundations of the CLT Emphasis Area: Instructional Design, Implementation, and Evaluation. July 1995. David Jonassen: Constructivist Learning Environments Gaynor C. Jeffery & Geoffrey Underwood: Combining Ideas in Written Text: The Role of Working Memory in the Development of a Writing Skill. Journal of Educational Psychology, March 1996, Volume 88, Number 1 Youngcook Jun, James Levin, Michael Jacobson: The Message Assistant: A communication tool for educational networks. Hunter Lawrence: Estimating Human Capacities (Size 1.0K) Patrick Mendelsohn: Les principaux paradigmes expérimentaux de l'étude des activités cognitives Renkl, A. (1995). Learning from worked-out examples: A study on interindividual differences (Research report No. 51). München: Ludwig-Maximilians-Universität, Lehrstuhl fόr Empirische Pädagogik und Pädagogische Psychologie. Stark, R., Gruber, H., Mandl, H. & Renkl, A. (1996). Wenn Expertise nichts nόtzt. Eine Replikationsstudie (Forschungsbericht Nr. 68). Mόnchen: Ludwig-Maximilians-Universitδt, Institut fόr Pδdagogische Psychologie und Empirische Pδdagogik, Lehrstuhl fόr Empirische Pδdagogik und Pδdagogische Psychologie. Sarah V. Stevenage: Can Caricatures Really Produce Distinctiveness Effects?
Sweller & Chandler 1991
Juhani E Tuovinen: Five Steps to Effective Learning of Computer Software by Reducing Cognitive Load Hans van der Meij: Visuele handleidingen: Ik zie ik zie wat jij niet ziet. Brent G. Wilson: Maintaining the Ties between Learning Theory and Instructional Design American Educational Research Association 1995. Brent G. Wilson, Peggy Cole: Cognitive teaching models In: Handbook of Research in Instructional Technology 1996. ED39: How We Learn Christine Alavi and Don Margetson: 1993 AARE conference - abstracts of papers and other presentations Psychology 300 Cognitive Psychology. Suggested Articles for Article Review and Independent Research Projects Bibliographie du Cours STAF 15 Partie 3: Recherche en Education et NTI : L'inventaire Bibliographie Bibliography of Additional Print Resources Department of Artificial Intelligence 1995: Publications School of Computer Science and Engineering 1995: Publications School of Computer Science and Engineering 1995: Publications School of Computer Science and Engineering 1995: Academic Units: Subject Descriptions: Faculty of Professional Studies: School of Education Studies School of Computer Science and Engineering 1995: Academic Units: Subject Descriptions: Faculty of Professional Studies: School of Education Studies School of Computer Science and Engineering 1995: Flash Information. Vol. 2 (1), August 22-26, 1994
|
|
|
Chris Mueller (prolingua@access.ch)
++41 (0)52 301 3301 phone
|
97 04 13 |
