Interfacing with Learning Technologies

1
Interfacing with Learning Technologies

• Shaaron Ainsworth

2
Overview

• Interface Issues in Learning Technologies
• Advantages of the right representation
• Properties of representations
• More than one representation (MultiMedia)
• Mayer
• Ainsworth
• Conclusions

3
Interface Issues in Learning Technologies

• For the first 30 years of computer-based
  learning, the interface was a non-issue.
• Early systems were all textual. As a result, the
  interface was almost a forgotten issue till
  around 1985.
• Usability of the Interface
• Interfaces should be design to be as easy to use
  and transparent as possible (see O'Malley 1990).
  (or should they?)
• Domain/Knowledge Representation
• How the knowledge is presented (e.g. what to put
  on a slide, should I use text or graphics,
  animations or static representations, etc)
• The focus of the presentation

4
Usability

• Human computer interaction The effectiveness,
  efficiency, and satisfaction with which specified
  users achieve specified goals in particular
  environments ISO 9241

5
Consequences of poor usability

• If you have the wrong interface, the environment
  may be sufficiently unusable that learners will
  find it difficult to use or may not to use the it
  at all.
• Or ..

6
Usability for Learning environments

• Learning to use a new system is not the same as
  learning though a system.
• Effects with technology or effects of technology
  (Salomon)
• Learning technologies may require different
  principles (Gilmore, 1996).
• For example, direct manipulation of an interface
  is not as good for learning the Towers of Hanoi
  as a command language interface

7
Overview

• Interface Issues in Learning Technologies
• Advantages of the right representation
• Properties of representations
• More than one representation (MultiMedia)
• Mayer
• Ainsworth
• Conclusions

8
Domain or Knowledge Representation

• What information do you chose to display to
  learners
• How do you chose to display it?
• What interactivity will you support?

9
Representations and Learning Technologies

• Now the interface is considered paramount
  receives a huge amount of attention during the
  design process.
• Computer-based Representations have a number of
  advantages
• Routine computations can be off-loaded
• Can focus learners attention on the essentials
  of the domain
• Representations can be placed under active
  control
• Interactive manipulation may help learners
  construct their own understanding of a domain
• Screen based representations may be more easily
  shared
• Multimedia - the "use of multiple forms of media
  in a presentation
• Consider a typical multi-media screen video,
  text, graphs, diagrams, spoken language...

10
A Typical Multi-Media Screen
11
But is this a good idea?

• Multimedia sells. But is it effective and how
  should it be designed?
• Two approaches
• Mayers Cognitive load theory of multimedia
  learning
• Ainsworths DeFT framework for learning with
  multiple ERs

12
Mayer Cognitive Theory of Multi-Media Learning

• Visual auditory experiences/information are
  processed through separate information processing
  "channels."
• Each channel is limited in its ability to process
  information.
• Processing is an active cognitive process
  designed to construct coherent mental
  representations

13
Typical Empirical Study

• Participants for an experiment are recruited in
  return for credit on psychology courses or are
  paid a small amount.
• What they learn does not relate to their
  education
• They may be given a short pen and paper
  multi-choice pre-test to check that they have
  little prior knowledge of the concepts of the
  domain and then are randomly assigned to two
  groups.
• A short orientation phase is provided to ensure
  that students know how to use the interface.
• They then learn for 30 minutes followed
  immediately by a pen and paper multi-choice
  post-test of the domain concepts, which typically
  will include some harder elements than the
  pre-test.
• They are debriefed, thanked for their
  participation and told not to sign up for further
  experiments, as they are not naïve to the
  material.
• The whole experience takes about an hour.

14
Multimedia

• From words and pictures than from words alone.
• Students who listened to a narration explaining
  how a bicycle tire pump works while viewing a
  corresponding animation gave twice as many useful
  solutions to transfer questions than did students
  who listened to the same narration without the
  animation (Mayer Anderson, 1991).
• Students build two different mental
  representations --a verbal model and a visual
  model -- and build connections between them.

15
Temporal Contiguity

• When corresponding words and pictures are
  presented simultaneously
• Mayer, Moreno, Boire Vagge, (1999) found that
  presenting simultaneous narration and large
  chunks of animation was better than sequential
  presentation
• Corresponding words and pictures must be in
  working memory at the same time in order to
  facilitate the construction of referential links
  between them

16
Principles Students learn more

• Multimedia From words and pictures than from
  words alone.
• Spatial contiguity When corresponding words and
  pictures are near each other
• Modality From animation narration than
  animation on-screen text.
• Coherence When extraneous information is
  excluded
• Temporal contiguity When corresponding words and
  pictures are presented simultaneously
• Redundancy From animation and narration than
  from animation, narration, and on-screen text.
• Individual Differences Effects are stronger for
  low-knowledge learners than for high-knowledge
  and for high spatial rather than from low spatial
  learners.

17
Mayer Analysis Positive

• Robust and replicable results confirmed by others
• Relationship between theory, design and
  evaluation
• Statistical rigour and experiments which explore
  conditions when multimedia is not effective
• Explore different forms of learning outcome (e.g.
  facts, transferable knowledge)
• The most popular current theory (see also
  Cognitive Load theory)

18
Mayer Analysis Minus

• Is the theoretical explanation sufficient? Are
  there other explanations equally consistent with
  the results?
• Is the methodology appropriate?
• Is the explanation sufficiently complete?
  Emphasis is placed on representation form and
  slightly on learning outcomes and individual
  factors but
• Are a sufficient variety of representations
  explored?
• Are a sufficient variety of learning tasks
  explored?
• Are most of the results about a specific form of
  dynamic representation?
• Is it too early for principles?

19
Overview

• Advantages of the right representation
• Properties of representations
• More than one representation (MultiMedia)
• Mayer
• Ainsworth
• Conclusions

20
An Alternative DeFT (Ainsworth, in press)

• In order to understand learning with multiple
  representations, we need to explore a wider
  variety of learning scenarios and provide a
  deeper account of the processes involved in
  learning
• Three key questions
• How is the system designed? (Design)
• What are you using the multiple representations
  for? (Functions)
• What cognitive tasks must learners perform?
  (Tasks)
• Ignores type of learning outcome (for now)

21
(No Transcript)
22
Cognitive Tasks

• the properties of the representation
• the relation between the representations and the
  domain
• how to select appropriate representations
• how to construct or even invent an appropriate
  representation
• how to translate between representations

23
The properties of the representation

• Learners must know how a representation encodes
  and presents information (the format).
• In the case of a graph, the format would be
  attributes such as lines, labels, and axes.
• They must also learn what the operators are for
  a given representation.
• For a graph, operators to be learnt include how
  to find the gradients of lines, maxima and
  minima, and intercepts.

24
The relation between the representations and the
domain

• Interpretation of representations is inherently
  contextualised
• It particularly difficult for learning, as this
  understanding must be forged upon incomplete
  domain knowledge.
• Learners need to determine which operators to
  apply to a representation to retrieve the
  relevant domain information
• For example, when attempting to read the velocity
  of an object from a distance-time graph, children
  often examine the height of line, rather than its
  gradient

25
How to select appropriate representations

• Learners may have to consider such aspects of the
  situation as the representation and task
  characteristics as well as individual
  preferences.
• Novick, Hurley Francis (1999) found that
  students were able to choose which of
  hierarchical, matrix or network representations
  was most appropriate to represent the structure
  of a story problem.
• But Cox (1996) found that learners without good
  insight into the problem tended to thrash about
  choosing representations without moving
  themselves nearer to a solution.
• Selecting appropriate representations will be
  more difficult for novices than experts as they
  can lack understanding of the deep nature of the
  tasks they are trying to solve (Chi, Feltovich,
  Glaser, 1981).

26
How to construct or invent an appropriate
representation

• Learners often construct their representations
  inaccurately (e.g. Cox 1996).
• However, learners could sometimes draw the
  correct inference even if they form incorrect
  representations.
• There is evidence that creating representations
  can lead to a better understanding of the
  situation. Grossen Carnine (1990) found that
  children learned to solve logic problems more
  effectively if they drew their responses to
  problems rather than selected a pre-drawn
  diagram.

27
How to translate between representations

• Learners find translating between representations
  difficult (e.g. Anzai, 1991).
• Learners can fail to notice regularities and
  discrepancies between representations (e.g.
  Dufour-Janvier, et al 1987).
• Teaching learners to coordinate MERs has also
  been found to be a far from trivial activity.
• Yerushalmy (1991) provided students with an
  intensive three-month course with
  multi-representational software that taught
  functions. In total, only 12 of students gave
  answers that involved both visual and numerical
  considerations and those who used two
  representations were just as error prone as those
  who used a single representation.

28
Cognitive Tasks Summary

• Many of the tasks that learners must undertake to
  learning with multiple representations are not
  trivial
• They multiply as more representations are used

29
Design Parameters
30
Ainsworth et al, (2002) Format
31
CENTS MERs Format Redundancy
Mixed
Maths
Picts
Categorical Magnitude
Continuous Magnitude Dir.
32
Results

• All experimental groups improved at estimating
• The maths/picts group improved at accuracy
  judgements, but the mixed group did not
• Use of the representations showed -

33
Can Learners Co-ordinate Representations?
0.8
0.6
Mixed
Correlations
Maths
0.4
Picts
0.2
Time 1
Time 2
34
DeFT analysis Positive

• In favour
• Considers a wider range of learning scenarios and
  takes more seriously the idea that different
  pedagogical functions require different
  multimedia designs
• Attempts a deeper analysis of cognitive processes
• Evaluations in more naturalistic situations
• Integrates a wider range of research

35
DeFT analysis Negative

• Says everything is more complicated!
• Has more questions than answers
• Still ignores learning outcomes
• Is less strongly related to a theoretical account
  of cognitive structure
• And like Mayer is not predictive

36
Conclusions

• Representations are crucial for learning. A
  learning environment must present information in
  such a way that it encourages learning.
• Learning with representations involves at least
  four factors Representation, Learner, Task and
  Outcome.
• Multiple representations/multi-media have an
  important role to play
• They may be motivating
• However, multimedia should be designed carefully
  to achieve the benefits without losing out to the
  costs
• How do so is still an open question..

37
Open Questions?

• How important is the interface in educational
  software?
• Is it more important to design for usability or
  learnability?
• What can the design of new interfaces learn from
  old interfaces? And do computers represent
  anything uniquely different?
• Do we know enough to design effective multimedia?
• Can classical cognitive psychological approaches
  explain learning with multimedia or do we need
  alternative perspectives?
• What new interface issues may arise in future?
• How should we evaluate the contribution that an
  interface makes to the success of an item of
  educational technology?

38
Reading From Course Text

• How People Learn
• Some relevant discussion in Chapter 3 and Chapter
  9

39
Reading From Original Sources

• Ainsworth, S.E. (1999) A Functional Taxonomy of
  Multiple Representations. Computers and Education
  
• Ainsworth, S. E., Bibby, P., Wood, D. (2002).
  Examining the effects of different multiple
  representational systems in learning primary
  mathematics. Journal of the Learning Sciences,
  11(1), 25-61.
• Ainsworth, S. E., Loizou, A. T. (2003). The
  effects of self-explaining when learning with
  text or diagrams. Cognitive Science, 27(4),
  669-681.
• Ainsworth (in press) DEFT a framework for
  learning with multiple representations on my
  website)
• Chi, M. T. H., Bassok, M., Lewis, M. W., Reimann,
  P., Glaser, R. (1989). Self-explanations How
  students study and use examples in learning to
  solve problems. Cognitive Science, 5, 145-182.
• Gilmore, D. J. (1996). The relevance of HCI
  guidelines for educational interfaces.
  Machine-Mediated Learning, 5(2), 119-133
• Larkin, J. H., Simon, H. A. (1987). Why a
  diagram is (sometimes) worth ten thousand words.
  Cognitive Science, 11, 65-99.

40
Reading From Original Sources

• Mayer, R. E., Moreno, R. (2002). Aids to
  computer-based multimedia learning. Learning and
  Instruction, 12(1), 107-119.
• Najjar, L. (1998). Principles of educational
  multi-media user interface design. Human Factors,
  40(2), 311-323.
• O'Malley, C. (1990). Interface issues for guided
  discovery learning environments. In M. Elsom-Cook
  (Eds.), Guided Discovery Tutoring London Paul
  Chapman Publishing
• Ploetzner, R., Fehse, E., Kneser, C., Spada, H.
  (1999). Learning to relate qualitative and
  quantitative problem representations in a
  model-based setting for collaborative problem
  solving. Journal of the Learning Sciences, 8(2),
  177-214.
• Scaife, M., Rogers, Y. (1996). External
  cognition how do graphical representations work?
  International Journal of Human-Computer Studies,
  45, 185-213.
• Seufert, T. (2003). Supporting coherence
  formation in learning from multiple
  representations. Learning and Instruction, 13(2),
  227-237..
• Verdi, M. P., Johnson, J. T., Stock, W. A.,
  Kulhavy, R. W., Whitman, P. (1997). Organized
  spatial displays and texts Effects of
  presentation order and display type on learning
  outcomes. Journal of Experimental Education,
  65(4), 303-317.
• Zhang, J., Norman, D. A. (1994).
  Representations in distributed cognitive tasks.
  Cognitive Science, 18, 87-122.

41
(No Transcript)
42
Route planning in London

• To travel between St Pancras and Victoria
• An Underground map shows you the stations and
  lines and does not preserve distance
• Walking tour of the London Parks
• An A to Z shows the streets, geographical
  features and real distance but does not tell
  you about the tube lines

Both
43
London Underground Map
44
A-Z Map
45
Shared information in a single representation
46
Constraining Interpretation

• A familiar representation can help you understand
  a more complex one

47
Deeper Understanding

• The Quadratic Tutor (Wood Wood, 1999)

48
Spatial Contiguity

• When corresponding words and pictures are near
  each other
• Students who read a text explaining how tire
  pumps work with captioned illustrations generated
  about 75 more useful solutions on transfer
  questions than did students who read the same
  text and illustrations presented on separate
  pages (Mayer, 1989)
• Corresponding words and pictures must be in
  working memory at the same time in order to
  facilitate the construction of referential links
  between them.

49
Modality

• From animation narration than animation
  on-screen text.
• Students who viewed a lightening animation with a
  narration generated 50 more useful solutions on
  a transfer test than the same animation with
  on-screen text (Mayer Moreno, 1998).
• On-screen text and animation overload the visual
  system whereas narration is processed in the
  verbal information processing system and
  animation is processed in the visual information
  processing system.

50
Coherence

• When extraneous information is excluded
• Moreno Mayer (2000) gave students a lightening
  animation with concurrent narration, or extra
  environmental sounds or with extra music, or all
  three.
• Music tended to hurt students' understanding but
  environmental sounds did not hurt
• Overload was created by adding unnecessary
  auditory material so fewer relevant words and
  sounds entered the cognitive system fewer
  cognitive resources was allocated to building
  connections amongst them.

51
Redundancy

• From animation and narration than from animation,
  narration, and on-screen text.
• Additional text harmed learners performance on
  transfer problems as this induced a
  split-attention effect

52
Individual Differences

• Effects are stronger for low-knowledge learners
  than for high-knowledge and for high spatial
  rather than from low spatial learners.
• Students with high prior knowledge may be able to
  generate their own mental images while listening
  to an animation or reading a verbal text so
  having a contiguous visual presentation is not
  needed.
• Students with high spatial ability are able to
  hold the visual image in visual working memory
  and thus are more likely to benefit from
  contiguous presentation of words and pictures.