What Happens First? The chicken or the (Easter) egg? - PowerPoint PPT Presentation

About This Presentation
Title:

What Happens First? The chicken or the (Easter) egg?

Description:

... The two faces of Gestalt psychology, American Psychologist, 41, pp. ... Kemp, S. (1987), Gestalt grouping effects in locating past events in timelines. ... – PowerPoint PPT presentation

Number of Views:85
Avg rating:3.0/5.0
Slides: 13
Provided by: radulu
Category:

less

Transcript and Presenter's Notes

Title: What Happens First? The chicken or the (Easter) egg?


1
What Happens First?The chicken or the (Easter)
egg?
  • Representation paradoxes in (and around) software
    design
  • Radu Luchianov
  • CogSci Research Seminar, Oct.24,2003

2
Chicken, then egg
  • Software is about solving problems by automating
    function
  • Notice problem that can be simplified
  • Find a good way to simplify it
  • Extend solution to other similar problems
  • Product reflects its function

3
Egg, then chicken
  • Concentrating on the solution, its easy to lose
    track of the problem
  • SE is taught in system-amenable domains
  • Theres more emphasis on system-system
    interaction than system-user feedback (e.g. file
    formats, standard interfaces)
  • Results
  • Product offers lots of functionality and poor
    ways to navigate or organize it
  • Users perceive less functionality in software
    tools than what actually is implemented (economic
    target?)

4
Paradoxes, conundrums and other pesky SE decision
points
  • Epistemic
  • Referential variety vs. formal tractability
  • Context dependence vs. strict control
  • Complexity and modularity vs. economy (Easter
    eggs)
  • Socio-Economic
  • Security vs. open knowledge
  • User orientation and feedback vs. privacy

5
Representation
  • Highly abstract, many interpretations,
  • What knowledge is made of.
  • Systemic Relationships of some sort among the
    elements of a specific domain or between two
    different domains.

6
Representation
  • Not necessarily tractable in symbolic terms
    (either discrete or continuous) (Arnheim, R.,
    1986).
  • Psycho-physiological Internalized predictors of
    past experiences, used in collating current (and
    future) experiences (Eklund, J., et.al., 1997,
    Kaptelinin, V., Bonnie, A. (1997)
  • Observations of consistent realms result in
    static representations (easily formalized, what
    SE modeling has been doing).
  • Less consistent realms require more dynamic
    representations (what SE modeling is constantly
    attempting, e.g. faster feedback, OOD function
    modularity, templating).

7
Predictors and collation
  • e.g. Rodgers, G.E., 1982
  • Psychological data shows that human memory is
    strongly influenced by the recall situation,
    which is why I call representations only
    predictors.
  • Past experiences serve in evaluating current
    experiences and automating response to them, thus
    collating.

8
Formal vs. guide
  • Communication requires a medium (thus records)
  • Strict representations are easier to record, but
    require lots of resources
  • (e.g. Formalization, Operationalization)
  • Loose representations lack precision, can lead to
    (guideline) conflict and make accountability
    difficult
  • Formal representation seems the way to go when
    resources are not a problem. (modern teaching vs.
    apprenticeship)

9
Domains
  • While learning skills (concrete-factual or
    procedural knowledge) in school-type settings,
    only very thin domains (toy domains), are
    presented however, in actual working
    environments, one has to apply knowledge in much
    broader domains. (Barr, A. Feigenbaum, E.A.,
    1981)
  • Broad SE domain including the epistemic facets
    as well as psycho-physiological, social and
    economic ones. (Luchianov, R., 2000)
  • Integration work (synthesis) is all about
    broadening the domain of analysis when
    integration is faulty, the resulting domain is
    fragmented, giving rise to cognitive errors (e.g.
    misinterpretations, paradoxes, lack of
    significant results, incorrect results)

10
What does all this mean?
  • Low reliability of knowledge-based experiments
    done outside of the environment where the
    knowledge is used
  • make use of field observation (Eklund, J. et.al.,
    1997), as well as an internally consistent model
    (e.g. Helson, H., 1964 Card, Moran Newell,
    1980 or Kaptelinin, V. and Bonnie, A., 1997)
  • use dynamic, situated lab observation (Nielsen,
    J., 1994),
  • whenever possible, choose apprenticeship over
    schooling (Zlate, M., 1999)
  • Self-deception A feeling of knowing more than
    one actually does know about a domain (Schacter
    Worling, 1985)

11
and in SE?
  • Emphasis on the user as part of the system
  • Standard starting points in the tools interfaces
  • Transparent functionality
  • Describe function, then tool use (e.g. stats)
  • Selectable levels of feedback (at use and
    creation time)
  • Personalized interaction (e.g. toolbars, menus,
    dialogs, automatic data display,
    context-sensitive tool display)
  • Which is the MonDoc way personalized tools in
    which the (artificial, economic) borderline
    between application and document is blurred.

12
References
  • Arnheim, R. (1986), The two faces of Gestalt
    psychology, American Psychologist, 41,
    pp.820-824.
  • Barr, A. Feigenbaum, E.A. (1981), The Handbook
    of Artificial Intelligence, vol.1, HeurisTech
    Press, Stanford, CA, USA. ISBN 0-86576-005-5
  • Brusilovsky, P., Eklund, J. (1999) A Study of
    User Model Based Link Annotation in Educational
    Hypermedia, Institute for Information Processing
    and Computer-related new Media, issue 4, vol.4,
    on-line athttp//www.iicm.edu/jucs_4_4/a_study_o
    f_user/paper.html
  • Card, Moran Newell (1980), The keystroke-level
    model for user performance time with interactive
    systems, in Communications of the ACM 23(7),
    pp.396-410
  • Eklund, J., Brusilovsky, P. Schwarz, E. (1997),
    Adaptive Textbooks on the World Wide Web,
    presented at the Third Australian World Wide Web
    Conference, 5-9 July 1997, Southern Cross
    University, Australia on-line athttp//ausweb.s
    cu.edu.au/proceedings/eklund/paper.html
  • Hadjiilieva, K., Kokinov, B., Yoveva, M. (1997).
    The influence of pragmatic distance on context
    effects. In B. Kokinov (Ed.), Perspectives on
    cognitive science, Vol. 3. Sofia NBU Press.
  • Helson, H. (1964), Adaptation-level theory,
    Harper Publishers, New York, NY, USA
  • Kaptelinin, V. and Bonnie, A. (1997), Activity
    theory Basic concepts and applications,
    Proceedings of CHI97 annual conference of the
    Association for Computing Machinery. on-line
    athttp//www.acm.org/sigchi/chi97/proceedings/tu
    torial/bn.htm
  • Kemp, S. (1987), Gestalt grouping effects in
    locating past events in timelines. Acta
    Psychologica, 64(2), 139-149.
  • Kokinov, B. (1997). A dynamic theory of implicit
    context. In Proceedings of the Second European
    Conference on Cognitive Science. Univ. of
    Manchester Press.
  • Lewis Polson (1995), Cognitive walkthrough, ACM
    archive, on-line athttp//www.acm.org/sigchi/chi
    95/Electronic/documents/tutors/jr_bdy.htm
  • Luchianov, R. (2000) COGS MoStaCon, on-line
    athttp//www.nbu.bg/cogs/personal/radu/proj/msc/
    index.html
  • Marks, M. (1999). A test of the impact of
    collective efficacy in Routine and Novel
    performance environments. In Human Performance
    12(3/4), pp.295-309, Lawrence Erlbaum Associates,
    Inc., Mahwah, NJ, USA. ISSN 0895-9285
  • Miller, G. A. (1956), The Psychological Review,
    vol.67, pp.81-97, on-line athttp//www.well.com/
    user/smalin/miller.html
  • Murdock, B.B. (1962), The serial position effect
    of free recall, Journal of Experimental
    Psychology, 62, pp.618-625.
  • Nielsen, J. (1994) Heuristic evaluation, on-line
    athttp//www.useit.com/papers/heuristic/heuristic
    _evaluation.html
  • Rodgers, J.E. (1982), The malleable memory of
    eyewitnesses, Science 82, 3(5), pp.32-35.
  • Scarr, S. Zanden, J.V. (1987), Understanding
    psychology, Random House, New York, NY, USA
  • Schacter Worling, (1985), Attribute information
    and the feeling-of-knowing. Canadian Journal of
    Psychology, 10, pp.379-399.
Write a Comment
User Comments (0)
About PowerShow.com