Bridging the Gap between Research
and Practice: Educational Psychologybased Instructional Design for
Developing Online Content
Stephanie T.L. Chu
Simon Fraser University
Burnaby BC, Canada
http://www.sfu.ca/~stephanc
(see “artefacts” section)
Research Context &
Purpose
Context
 Significant research has been performed in Educational
Psychology.
 Fields such as human-computer interaction, graphic arts
and media design have had more influence in developing
online materials.
 Instructional designers may not be aware of principles (and
rationales) which stem from Educational Psychology
research.
 A gap exists between research findings and real-life
application.
Purpose
To aid content developers/instructors in making instructionally
sound decisions when designing learning materials.
This Research…
Aims to
 Perform a comprehensive review of the past 25 years of
Educational Psychology research literature.
 Raise awareness about fundamental considerations.
 Derive a rich set of instructional design guidelines for
developing learning materials.
 Research  ID principles  Content development.
 Identify categories
 Text presentation, examples/problems, multimedia...
 Include specific conditions where techniques will or will not
work, identify relationships and conflicts between
techniques.
Note: Select findings from the past five years are covered in this
paper and presentation.
Overview
Learning
Cognitive Load
Theory
Learner
Experience
Instruction
and Content
Cognitive Load Theory
Working Memory
Schemas
Cognitive Capacity
Controlled
Intrinsic Load
Automated
Germane Load
Extraneous
Load
Cognitive Skills Acquisition
Working Memory
Schemas
Cognitive Capacity
Intrinsic Load
Controlled
Automated
Early
Intermediate
Late
Germane Load
Extraneous
Load
Cognitive Load Theory
Working Memory
Auditory/Verbal
Channels
Schemas
Controlled
Automated
Visual/Pictorial
Cognitive Capacity
Intrinsic Load
Germane Load
Extraneous Load
Theoretical Implications

Consider the mental load imposed by the learning
material, learning processes, and unrelated activities.


Example: Making “full use” of a computer-based medium vs.
computers as a text-repository.
Make use of dual-modality instruction: Present material in
verbal and pictorial form concurrently. (Multimedia Principle,
Moreno & Mayer, 1999)


Example: Animated processes with images and text.
Consider appropriate conditions. For example, avoid
asking learners to:
 Integrate mentally corresponding representations which
are separated in space or time (Split-source Instructions,
Kalyuga, Ayres, Chandler & Sweller, 2003).

Engage heavily in cross-referencing and search-andmatch processes.
The Learner: Learner
Experience
Recent research investigates interactions
between levels of learner knowledge
(experience/expertise) in a domain and
levels of instructional guidance.


Additional guidance may increase or decrease
cognitive load.
Expertise affects whether an information source
should be eliminated or not.

Novices may need redundant information, but more
experienced learners may experience increased
extraneous load.
Note: See Kalyuga, et al. (2003) for split-attention effect, redundancy effect, expertise
reversal effect, elimination effect, and guidance-fading effect.
Learner Experience &
Computer-based Learning

Design instruction that
supports the learning process and
 adapts instructional procedures and
techniques to support individual
learners.



Build in prior knowledge and formative
assessments (so the system can adapt).
Enable some user control over multimedia
such as the ability to eliminate one
information source or control its pace.
Instruction and Content:
Categories

*Problems


*Multimedia


Problem solving, worked examples.
Images and animation; animation and
narration; pedagogical agents; “when
less is more”; sequencing;
Textual content

Placement of main ideas and
examples; text cues; text and pictures;
maps
Problem Solving, Worked
Examples & Problems


Clearer definition of “examples” and
“problems” and conditions.
Relationship between worked
examples/problems and learner experience
and cognitive load


Worked examples are effective during
initial stages of cognitive skill
acquisition, while in later stages,
solving problems is superior.
Fading out: gradual elimination of worked
examples (Renkl, Atkinson, & Maier, 2000).
Mapping out
Learning &
Experience
Against the
Use of
Examples &
Problems
Early
Intermediate
Late
Experience
Little or no domain knowledge
Gaining domain knowledge
More/high domain knowledge
Learning
Basic understanding
Start to solve problems (understand
domain & apply)
Actual problem solving; optimization
(understanding acquired)
Task
Study materials
Use abstract principles to solve concrete
problems
Practice (speed & accuracy)
Approach
Worked examples
Worked examples for learners to actively
self-explain/reflect. Move towards
increased use of problems to solve
to develop skill in this.
Anticipating and imagining a previously
learned solution path (introduction
to problem-solving elements) for
higher prior knowledge learners
Problem solving (reversed worked
examples effect due to
redundancy effect)
Skill
Gain basic understanding of
domain
Develop ability to generalize over surface
structures
Proceed in skill acquisition
Automation of at least a
subcomponent of skills
(speed and accuracy)
Intrinsic
load
The material being studied
Material being studied (gradual decrease in
load as cognitive skill acquisition
increases)
Material being studied (less load)
Germane
load
Worked examples
Self-explanations (principle-based
explanations, goal-operator
relations, coherence among
examples)
Anticipating and imaging based on what
has been learned to-date
Problem-solving
Extraneous
load
Steps in problem solving
Steps in problem solving
Redundant information (examples)
Self-explanations
Alternative
(Renkl & Atkinson, 2003)
Fading worked-out solution steps:
1.
Present concrete example (model)
2.
Present example where one single solution step is omitted (coached problem
solving)
3.
Increase the number of blanks step-by-step until only the to-be-solved
problem remains (independent problem solving)
N/A
Cognitive
load
N/A
Fading reduces heavy cognitive load and reduces errors in learning. By
implementing a step-by-step approach, demands gradually increase.
Appropriate
for
N/A
For problems solvable by applying specific “to-be-learned” rules (near-transfer)
where reducing errors is advantageous
Fading can be used for problems that require modification of learned solution
methods (far transfer). Though learned roles cannot be directly applied,
errors may trigger reflection and deepen understanding of the domain.
Problem Solving, Worked
Examples & Problems

Presenting content online may provide
options for students as their experience
levels increase


System moves learners from worked
examples to problems or “fade-out”.
Multimedia can offer novel instructional
opportunities:
Learning objects can depict an
expert’s thought processes.
 Aural and visual modeling.

Multimedia

Numerous studies examine using both channels of
working memory such as through animation. (Mayer &
Associates (i.e. Moreno, 1998-present) and others (e.g. Atkinson,
2002))

Key ideas:
 Learning materials that make use of both the visual and
verbal channels of information processing (working
memory) are more effective for learning than a single
representation of information or multiple
representations through the same modality.
• Example: Animation and narration rather than narration
alone.
• Example: Animation and narration is a better combination
than animation and on-screen text (both require
processing by visual channel).

Animation and narration should be presented
concurrently rather than consecutively so that students
develop a stronger sense of the association between
the information seen and heard.
Multimedia


Research on pedagogical agents (animated “guides”) is
fairly new and requires further examination.
“When less is more”



More ways of presenting the same information within
resources (information delivery hypothesis), is not
necessarily better (cognitive load effect, redundancy effect)
For optimal multimedia use, select resources that have only
one verbal and visual component (avoid split-attention
effect).
Seductive details (“interesting” unimportant information
related to the learning materials) may have neutral to
negative effects.
• Thought to encourage interest in the topic (emotional interest
hypothesis), but this is not supported.

Design or select materials that do not contain extraneous
words, sounds (including music), and video which can
distract learners (take away from germane cognitive load).
In Conclusion
General Implications



Need for designers to consider cognitive processes
associated with learning.
The extent to which cognitive load comes into play
depends on the learner (specifically, domain
experience/expertise).
Computer-based learning can provide novel interactive
learning opportunities which capitalize on multimedia and
the ability to adapt to an individual learner.
Paper and presentation available:
 http://www.sfu.ca/~stephanc
 (see “artefacts” section)