Chapter 3: Attention in Perception and Display Space

1
Chapter 3 Attention in Perception and Display
Space

• Captain Jameel Adam
• DRDC Toronto
• Operational Human Engineering Group

2
Attention (1/3)

• Selective Attention
• Ability to select which perceptual channel to
  focus attention on
• e.g., visual sampling, target search
• Influences design of
• Instrument layout
• Operating procedures
• Search tasks etc.

3
Attention (2/3)

• Divided Attention
• Ability to divide attention simultaneously
  across two or more perceptual channels
• e.g., timesharing, display monitoring
• Influences design of
• Integrated displays
• Operating procedures
• Multitasking situations etc.

4
Attention (3/3)

• Focused Attention
• Ability to focus on a particular perceptual
  channel and exclude other (adjacent) channels
• Opposite of divided attention
• e.g., target search, studying, translating
• Influences design of
• Work environments
• Warning systems etc.

5
Spotlight Metaphor

• We attend where beam falls
• Information sometimes outside of beam we should
  attend to it (Selective)
• Other irrelevant information in beam, need to
  narrow beam (Focused)
• When we need to get information from various
  sources, widen the beam (Divided)

6
Selective Attention

• How do we sample various sources of information?
• Aviation research(T layout)

7
Selective Attention

• 2 types of Visual Sampling
• Supervisory Control Sampling
• Operator scans display of a complex system in
  order to control it (power plant, aircraft)
• Location known, information unknown
• Target Search Sampling
• Operator trying to locate target (radar, search
  and rescue)
• Information known, location unknown

8
Supervisory Control Sampling

• Based on optimality
• Maximizing expected return for sampling
• Sampling guided by underlying model of system
  (can also indicate understanding of model)
• Sampling adjusted to event rate (sluggish ?)
• Frequent events sampled more, vice versa

9
Supervisory Control Sampling

• Samples done more often than required due to
  imperfect memory
• Sampling costs reduce rate (digging for
  information use reminders)
• Sampling affected by arrangement (diagonal less
  likely than horizontal)
• Preview of upcoming events can improve
  performance (but working memory limits this)

10
Supervisory Control Sampling

• Sampling strategies can be affected by
  arousal/stress
• Cognitive Tunnelling
• After problem, fixation locked on one indicator
  until returns to normal

11
Visual/Target Search

• Based on expectancies
• Novice vs. expert
• Display factors and salience
• Large, bright, colourful, dynamic ? attract
  attention
• Physical location (left to right, centre first
  etc.)

12
Visual Fixations

• Centre of fixation (approx 2 deg)
• Useful Field of View (UFOV)
• Effective area within which information can be
  extracted
• Varies between 1 and 4 deg
• Dwell time
• Length of fixation

13
UFOV

• Affected by
• Density of information
• Discriminability of target from background
• UFOV doesnt blanket area
• Targets may not be identified in UFOV
• Decision criterion, training

14
Dwell Time

• Longer dwells on
• Less legible displays
• Less familiar words
• More difficult text
• Unusual objects, out of context

15
Factors Affecting Search Speed/Time

• Number of items
• Typically serial search
• Target distinguished by single dimension/simple
  rule
• Typically a parallel search
• Discriminability
• Hard to distinguish ? serial
• Easy to distinguish ? parallel

16
Factors Affecting Search Speed/Time

• Absence/presence of target features
• Absent (O in Qs) ? serial
• Present (Q in Os) ? parallel
• Spacing
• Little effect scanning and clutter trade off
• One or more targets
• Generally less time if single type of target
  (more mental templates)
• Unless all targets discriminable along single
  dimension (e.g., diagonal stroke in K, X)

17
Factors Affecting Search Speed/Time

• Training
• If one can train to automaticity, leads to
  parallel processing
• Only where targets are consistent, must maintain
  mapping

18
Parallel Processing and Divided Attention

• Preattentive phase
• Automatic
• Organizes visual world into objects and groups
• Short term sensory store
• Subsequent selective attention
• Further elaboration
• Perception

19
(No Transcript)
20
Global and Local Processing

• Global processing
• All items of organized display processed together
  in parallel
• Local processing
• Processing single object in display

21
(No Transcript)
22
Gestalt Principles of Display Organization

• Factors that facilitate pre-attentive grouping of
  objects include
• Proximity
• Similarity
• Common fate
• Closure
• Movement
• Colour

23
Emergent Features

• Global processing often associated with features
  not present in the individual elements
  themselves, that emerge put together into a set
• Only works if the organization is compatible with
  task and Gestalt principles upheld

24
Spatial Proximity

• Expect spatial proximity will support parallel
  processing ? divided attention.
• However, spatial proximity does not guarantee
  that parallel processing will occur

25
Heads-Up Displays (HUDs)

• Idea is to superimpose aircraft instrument
  information on view of real world
• Ideally both can be attended in parallel, or at
  least reduce transition time

26
Break

• 15 Minutes

27
Spatial Proximity Conflicts

• Penalty involved in putting information close
  together
• If two attentional channels are close together,
  they will both be processed, even if only one is
  desired

28
Eriksen Eriksen

• Control
• H right button
• F Left button
• Display Clutter
• K H K

• Redundancy Gain
• H H H
• Response Conflict
• F H F

29
Eriksen Eriksen

• Display clutter
• Confusion, failure to focus between relevant and
  irrelevant
• Redundancy gain/Response conflict
• Proximity allows two channels to be processed in
  parallel, even if undesired
• Information competes at perceptual level
• If they agree, improved performance ? Redundancy
  gain
• If they disagree, reduced performance (?RT and/or
  errors) ? Response conflict

30
Object Based Proximity

• Similarity can be based on more than proximity
• Colour, shape, location, size, etc.)
• Stroop Effect
• GREEN RED

31
Object Displays

• Take advantage of object based proximity
• Integrating information sources into one stimulus
  if parallel processing required
• Integral dimensions can sometimes be achieved by
  combining dimensions into a single stimulus
• Emergent features property of whole, not seen
  when parts in isolation

32
Proximity Compatibility Principle

• If task requires HIGH processing proximity, then
  should have HIGH display proximity
• If task requires LOW processing proximity, then
  should have LOW display proximity
• Processing proximity extent to which
  information sources are use in the task
• Display proximity how close display elements are

33
Focussed Attention Tasks?

• What if the task requires focussed attention,
  will the single object interfere?
• Yes
• Use appropriate object type for task(s)

34
Using Multiple Sources of Information

• Integrating multiple information sources by
  proximity or common representation (objectness,
  colour, etc.) results in
• Increased probability of parallel processing
• The possibility of emergent features with
  improved performance if they are mapped to the
  task
• Possible creation of clutter or response conflict
  if the task requires focussed attention on one
  dimension

35
Designing Tasks Involving Divided Attention

• Minimise number of potential sources of
  information
• If timesharing is necessary, indicate, where
  possible, the relative priorities of
  tasks/stimuli
• Minimise difficulty level of tasks, where possible

36
Designing Tasks Involving Focussed Attention

• Make competing channels as distinct from each
  other as possible
• Separate competing channels physically as far
  apart as possible
• Minimize the number of competing channels
• Maximize salience of the most important
  channel(s), using colour, brightness, loudness,
  central location, etc.

37
Attention in Auditory Domain

• Does not follow searchlight model of visual
  domain
• No earballs to move around

38
Shadowing Technique

• Subject wears earphones with different message
  presented to each ear
• Required to shadow one message
• Repeat that message out loud

39
Results

• Can pick up some info from unattended channel,
  but has to differ
• Tone versus voice
• Male vs. female pitch
• Large change in volume

40
Results

• Report by order of ear, not time
• L 8 drive 7
• R leak 6 dog
• Subject 8, drive, 7, leak, 6, dog
• (i.e., LLLRRR not LRLRLR)

41
Broadbent Model

• Selective filter used to attend to one channel
• Switching operate on physical characteristics of
  message
• Loudness, pitch (male/female), quality of sound,
  etc.
• Problem cocktail party effect
• Even when not different (pitch, volume) from
  other message, subjects attended to their name

42
Triesman Model

• Semantic differences cocktail party effect
• Have a lexicon or mental dictionary of words
• When lexical entry fires, word is recognized
• Two factors
• Strength of signal (attended channel vs
  unattended)
• Context (threshold for lexicon reduced by common
  words that occur around it football game
  played versus football game experiment)

43
Implications

• Physical differences between channels are
  detectable
• Could make one louder but increases stress
• Leads to non-optimal performance
• Can use tendency to switch to contextually
  relevant information
• e.g., label the upcoming information so operator
  knows it is relevant

44
Implications

• Pilot asks for cloud cover info.....response
  starts with Cloud cover over....
• Or use personalized alerts Bill you are low on
  gas
• Context is important

45
Enough Theory

• What about real world applications?

46
Discussion - Land Warrior HMD

• Is this a good idea?
• Task types?
• Focussed/Divided attention?

47
Discussion 3D Positional Audio

• Will it work?
• Task types?
• Focussed/Divided attention?

Here
Here
Where are you?
Here