Interfacing with Learning Technologies - PowerPoint PPT Presentation

1 / 52
About This Presentation
Title:

Interfacing with Learning Technologies

Description:

Chi, M. T. H., Bassok, M., Lewis, M. W., Reimann, P., & Glaser, R. (1989) ... Machine-Mediated Learning, 5(2), 119-133. Larkin, J. H., & Simon, H. A. (1987) ... – PowerPoint PPT presentation

Number of Views:22
Avg rating:3.0/5.0
Slides: 53
Provided by: crippscomp
Category:

less

Transcript and Presenter's Notes

Title: 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.
Write a Comment
User Comments (0)
About PowerShow.com