Cognitive Systems Engineering Center

1
Cognitive Systems Engineering Center
Bridging the gap between CWA and application
design? Cognitive Systems Engineering Workshop
Seattle, November 2004 Annelise Mark
Pejtersen Cognitive Systems Engineering
Center Risø National Laboratory NORDICHI
Conference Workshop on Bridging the gap between
field studies and design

2
Cognitive Systems Engineering Center

• What does CSE offer to help?
• A theory of human adaptation and goal directed
  behavior
• Closed loop feed back system
• A generic framework for CWA and Evaluation
• 5 Design maps link CWA and interface design
• Design principles are embedded in theory and
  design maps and implications need to be extracted
  by the analyst from his expertise on the
  particular domain
• Design principles are domain dependent

3
Cognitive Systems Engineering Center
Theory....humans in dynamic work Adaption Activit
y under varying conditions will depend on
continous adaptation and improvisation, on the
ability to reconfigure patterns of behavior, to
modify effective routines, to combine elementary
routines into new patterns and to generate new
work processes on demand Constraints Constraints
and human performance criteria will define the
boundaries of the space in which the actors can
navigate freely according the their individual
resources and subjective performance
criteria Design principle Design a constraints
Space within which actors can improvise in a
particular situation that cannot be foreseen


4
Cognitive Systems Engineering Center
What action possibilities given the constraints
in this situation?

Means ends Information serves as criteria for
choice of action

Functional relationships?Multiple Constraints?
Are they invariant? Can they be negotiated?
5
Cognitive Systems Engineering Center
Design map 1 Domains, Tasks, Users Design map
2 Information content Design map 3
Navigation . Design map 4 Interface content and
form (Means ends and Skill-rules-knowledge based
behavior) Design map 5 Visual decomposition
display Design map no? Bridge the gap between
CWA and the functionality of the technical
system Each design map illustrates design
principles Empirical evaluation experiments and
user participation


6
Cognitive Systems Engineering Center
Analysis of Human-Work Interaction

Why? What? How?
Means ends is the backbone of field studies,
design, evaluation
7
Cognitive Systems Engineering Center

Brief review of design maps intended to help
bridge work analysis and design Examples Process
control stable task, dynamic process Library,engi
neering design, arhives, dynamic, diverse tasks,
create new information
8
Cognitive Systems Engineering Center
Design Map 1 Domain-Task-Users of the System


Engineering design, Archives, Libraries Loosely
coupled, collaborating autonomous users
9
Cognitive Systems Engineering Center
Design map no? Bridge the gap between CWA and
technical system design


Map Made in 80ties for library system
10
Cognitive Systems Engineering Center
Design map 2 Information Content


Means ends-Part whole
11
Cognitive Systems Engineering Center
Design Map 2 Information content of books

Means Ends levels in Database

12
Cognitive Systems Engineering Center
Design Map 3 Navigation in Information Content
by Means-Ends levels


13
Map 3 Navigation in Information Content by Task,
Strategies and Metaphor


14
Design Map 4 Interface Content and Form


15
Design Map 4 Interface Content and Form


16
Cognitive Systems Engineering Center
Design Map 5 Visual composition of display


Means Ends, Part Whole
17
Cognitive Systems Engineering Center

Design Map 5 Visual composition of display

Means Ends, Part Whole
18


19
Cognitive Systems Engineering Center

• Question you are dying to discuss
• CSE helps in Bridging gap between CWA and design?
• Generic, too general, and domain independent?
• Field study is rich source for ideas in the
  creative design process?
• Compatible frameworks support bridging the gap by
  empirical evaluation experiments with users?
• CSE is a skeleton and means ends the back
  bone?
• Implications of skeleton for design have to be
  deduced by the analysist from domain studies?
• Skeleton Muscles functionality and form must be
  created by the designer from domain specific
  studies, evaluations, other
• contextual factors?

20
Cognitive Systems Engineering Center
General enables comparison of constrints in
several domains to design a collaboratory

Goal to mediate national film culture- preserve
culture Scientific paradigmsPositivism,
Hermeneutic, Pluralism Management
styleAutocratic, democratic consensus,
chaotic ad hoc Work organisationHierarchi
cal, project organised, selforganised driven
by user service Common tasks and decisions,
analysis, interpretation, comparison, and
materials Different collaboration technology

Goals and constraints Priority measures Tasks Work
processes Physical objects
21
Cognitive Systems Engineering Center

What action possibilities given the constraints
in this situation?

Means ends Information serves as criteria for
choice of action

Functional relationships?Multiple Constraints?
Are they invariant? Can they be negotiated?
22

• Design principles
• Information about constraints must be available
  to be used as criteria to select action that is
  most effective
• Make constraints visible and easily perceptible
  in interfaces for different kinds of cognitive
  behavior
• Knowledge based,Rule-based and Skill based
  Behavior
• What does CSE offer to help design?
• Frameworks of CWA, Interface design and
  Evaluation bridge the gap

23

• autonomous users
• actors are actively formulating objectives and
  constraints
• many constraints are negotiable and are changed
• actors with different disciplines collaborate in
  decision making diverse domain information
  increases rapidly
• dynamic, diverse tasks that include creation,
  interpretation, negotiation and sharing of new
  knowledge
• ators organise their work and collaboration ad
  hoc and need to know each others objectives.
• Tightly coupled dynamic domain process, stable
  task
• Loosely coupled stable domain process, dynamic
  task

Cognitive Systems Engineering Center