Things That Make Us Smart Cognitive Artifacts

1
Things That Make Us Smart Cognitive Artifacts

• Course Notes
• Kathryn Summers
• 2009

2
Technology

• Cognitive artifacts can extend human cognitive
  power in these ways
• Reduce memory load
• Increase access to information
• Facilitate collaboration/communication
• Cognition??cognitive artifacts

3
Technology facilitating tasks

• Provide mental aids, reduce demands on knowledge
  in the head
• Use technology to make things more visible and to
  provide better feedback
• Use technology to automate the task, so the user
  doesnt have to do it
• Use technology to change the nature of the task,
  to make it easier

4
Cognitive artifacts--Changing the task

• Personal point of viewchanges the task
• Writing changes task from memory to
  writing/keeping track of/reading
• System point of viewpersonartifact is smarter
  than either one alone

5
Technology disadvantages

• Can make us LESS smart
• Can reinforce power disparities, economic
  disparities, even informational disparities (the
  digital divide)
• Can force us to behave like machines (science
  finds, industry applies, man conforms)

6
Normans law

• Technology should complement human abilities
• Compensate for weaknesses
• Enhance strengths
• Grudins lawtechnology will fail or be subverted
  when those who benefit are not those who do the
  work

7
Balancing design priorities

• User goals and desires
• What do users need to do
• What do users want to do
• Branding goals, business goals
• Technological constraints, cost constraints,
  physical manufacturing constraints

8
Why designers are not users

• Designers are experts about the system
• When they use the system, they mostly work from
  knowledge in the head
• Users are experts about the task
• Beginning users have to work from knowledge in
  the world to use the system
• Not all users are alike (physically, culturally,
  educationally, etc.)

9
Natural or evolutionary design

• Each design is studied
• Incremental changes made
• Changes evaluated
• New incremental changes made
• Drawbacksits slow, and requires flexibility,
  attention, and commitment

10
The design process (from Donald Norman)
Designers mental model
Users mental model
Translate mental model into design decisions
Interpret perception
Perceive system image
Build system image
System image
11
What to make visible (or audible)

• What can be manipulated? What kind of
  manipulation is possible?
• What is currently happening?
• What options are available?
• What are the results of the users action
  (feedback)?

12
Memory (knowledge in head)

• Arbitrary memorylimited to 5-7 things at most
• Memory for meaningful relationshipsfruit of good
  mappings
• Memory through explanationfruit of a good mental
  model

13
Knowledge in world

• Serves as its own reminderthe signal, and the
  message
• Ben Shneidermandirect manipulation interfaces

14
Using constraints

• Physical constraints (only five spaces for zip)
• Semantic constraints (rider needs to see)
• Cultural constraints (words go right to left)
• Logical constraints (one piece left)

15
Representation

• Abstraction/ representation enable cognition by
  removing perceptions/experiences from some of
  their details and allowing them to be manipulated
• A good representation captures essential elements
  and leaves out the rest a misleading
  representation can lead to faulty thinking,
  faulty conclusions
• Which features are relevant depends on the task
  and the goal.
• We value what we can represent. Things not
  represented get forgotten or diminished.
• Solving a problem means representing it so that
  the solution is obvious. A good representation
  may turn a (hard) reflective task into an (easy)
  experiential taske.g., tic-tac-toe

16
Information display

• Find the relevant information
• Process the information to generate the desired
  conclusion
• Information display solutions need to support
  the needs of all interested parties (e.g.,
  prescriptionsshould meet the needs of doctors,
  nurses, pharmacists, patients see matrix
  solution on pg 65)

17
Additive or substitutive

• Additive representations (tally marks, intensity)
• Substitutive representations (arabic numerals,
  hue)
• Examplebudgeting program for LD users

18
Error prevention

• Use natural and artificial constraints. Add
  forcing functions (where necessary for
  safetyavoid annoying users)
• Make options and results visibleprovide
  information in the world, reduce reliance on
  memory, or knowledge in the head
• Keep decision structure for key tasks either
  shallow or narrow
• Think of errors as attempts to do the task
  require less accuracy

19
Error recovery

• Minimize consequences for error
• Make actions reversible
• Make errors easy to discover
• Be polite

20
Learning how to use artifacts

• In the past, pieces of artifacts were generally
  physically visible smart folks could figure it
  out by lookingsurface artifacts
• We interpret artifacts based on their
  affordances, mappings, and constraints
• Now the relationships between controls,
  indicators, and system state are arbitrary and
  can be invisible (ergo, good design makes things
  visible)internal artifacts, in need of an
  interface
• (for examplepeople are internal artifacts)

21
Cognitive artifactslides for presentation

• Shared workspacewe can all see/analyze the
  points at the same time, we are all free to add
  new insights (communication, collaboration)
• External memoryincreasing memory permanence and
  memory quantity
• Spatial layout of slidesallows perceptual
  processing, helps point out relationships between
  ideas
• Multiple delivery modesallows people to focus on
  what they see OR on what they hear (may support
  individual learning styles)

22
A good system image will . . .

• Encourage users to explore a system and enable
  them to form a good mental model by
• Making things visible, including options and
  feedback
• preventing errors or making them easily
  recoverable
• The result is users who can use the system
  effectively, and who will probably enjoy their
  experience.

23
Thinking about learning cognition
24
Two kinds of cognition

• Experiential
• We perceive events and react efficiently and
  effortlessly
• The mode of expert behavior and flow
• Typically involves active participation
• Reflective
• We compare, analyze, and make decisions.
• The mode of new ideas and novel responses

25
Learning and cognition

• Learninga relatively permanent change in
  behavior potential due to practice or experience
• Cognitionthe processes by which sensory input is
  transformed, reduced, elaborated, stored,
  recovered, used (Ulric Neisser, 1967)

26
Observational learning

• Attentionidentify relevant info, pay attention
• Encodingprocessing the info
• Retentionstorage and retrieval
• Emissiondoing the behavior yourself

27
Multimedia and learning

• Norman worries that multimedia environments lean
  naturally towards experiential cognition,
  interfering with reflection.
• Multimedia environments are likely to involve
  event-driven processing or pattern-driven
  processing (recognition rather than analysis)

28
Experiential vs. Reflective tools

• Experiential cognitive toolsallow us to
  experience and act on the world (telescopes)
• Should provide lots of sensory stimulation
• Should NOT require reflection, analysis, or
  problem-solving to use
• Reflective cognitive toolsallow us to modify and
  act on representations
• Should support comparison, exploration,
  problem-solving
• Should not overwhelm the attention or restrict
  attention to a tiny piece of the information

29
Cognitive Processes in Learning

• Accretionaccumulation of facts
• Tuningpractice
• (5000 hrs, 2 full-time yrs to become an expert)
• Restructuringreflectively forming the right
  conceptual structure, changing how you understand
  the activity
• Flowabsolute absorption in the activity (can be
  disrupted by the tool)

30

• Informal learning
• Unstructured
• Group activity
• Goal is motivated
• Fun
• No interruptions
• Frequent flow experiences
• Self-pacing
• Topics, time, place are freely chosen
• Participants can be any age

• School learning
• Structured
• Individual activity
• Goal not motivated
• Fun not relevant
• Interruptions
• No flow experiences
• Pacing is fixed, forced
• Topics, time, place fixed
• Participants between ages of 6-20