L3- Divided attention
Dividing attention between stimuli
Selecting + keeping track of multiple stimuli
Change blindness in a natural scene
Campus change blindness experiment
- The findings showed that 50% did NOT notice the person change
Inattentional blindness
Observers monitoring an event in a visual display typically fail to notice an unexpected
visual event happening in the display they are attending to
E.g. Simons + Chabris (1999) had observers watch a video of an in-door ball game by two
teams marked by the colour of their shirts- white + black
- When observers were instructed to monitor the number of ball passes between the
members of the white team- 50% failed to notice the unexpected event of a person
dressed as a gorilla walking in between the players
- Detection of the unexpected event increased to around 70% when attention was
drawn to the black team
Tracking moving objects
Multiple object tracking (MOT)- observer
selectively tracks only the marked objects
among visually identical distractors
The target (to-track) items are visually
identical to the distractor (to-ignore) items- they are
shown here as distinct only for illustration
How many items can we track?
- All else being similar- we can track accurately
about 4 targets
- The more targets- the less efficiently we can
track them
- HOWEVER- this limit is NOT absolute
Density- packing the items together impairs tracking
accuracy
Location in the visual field- tracking 2 targets is easier if they
each are moving within different hemi fields (halves) of the
screen (visual field)
Speed (velocity)- tracking fast-moving targets is more difficult
, Multi-object tracking is object-based
Tracking elements have to be perceived by the visual system
Multi-object tracking is guided by ‘indices’ or ‘tags’ to objects
- This is to cope with continuous feature changes
A new metaphor for attention
Spotlight
- Attention moves in the visual field like a spotlight
- Spotlight has fixed size
- Selection is guided by pre-attentive processing
Zoom lens
- Attentional spotlight with variable size
Serial/parallel processing
Single-feature search- shallow search slopes (<10 msec/item)
Conjunction search- steep search slopes (>>10 msec/item)
Interpreting search slopes
Large variability in conjunction
search slopes
Not simply serial/parallel- rather
talk of efficiency
Hybrid models of stimulus
processing in visual search
Dividing attention between stimuli
Selecting + keeping track of multiple stimuli
Change blindness in a natural scene
Campus change blindness experiment
- The findings showed that 50% did NOT notice the person change
Inattentional blindness
Observers monitoring an event in a visual display typically fail to notice an unexpected
visual event happening in the display they are attending to
E.g. Simons + Chabris (1999) had observers watch a video of an in-door ball game by two
teams marked by the colour of their shirts- white + black
- When observers were instructed to monitor the number of ball passes between the
members of the white team- 50% failed to notice the unexpected event of a person
dressed as a gorilla walking in between the players
- Detection of the unexpected event increased to around 70% when attention was
drawn to the black team
Tracking moving objects
Multiple object tracking (MOT)- observer
selectively tracks only the marked objects
among visually identical distractors
The target (to-track) items are visually
identical to the distractor (to-ignore) items- they are
shown here as distinct only for illustration
How many items can we track?
- All else being similar- we can track accurately
about 4 targets
- The more targets- the less efficiently we can
track them
- HOWEVER- this limit is NOT absolute
Density- packing the items together impairs tracking
accuracy
Location in the visual field- tracking 2 targets is easier if they
each are moving within different hemi fields (halves) of the
screen (visual field)
Speed (velocity)- tracking fast-moving targets is more difficult
, Multi-object tracking is object-based
Tracking elements have to be perceived by the visual system
Multi-object tracking is guided by ‘indices’ or ‘tags’ to objects
- This is to cope with continuous feature changes
A new metaphor for attention
Spotlight
- Attention moves in the visual field like a spotlight
- Spotlight has fixed size
- Selection is guided by pre-attentive processing
Zoom lens
- Attentional spotlight with variable size
Serial/parallel processing
Single-feature search- shallow search slopes (<10 msec/item)
Conjunction search- steep search slopes (>>10 msec/item)
Interpreting search slopes
Large variability in conjunction
search slopes
Not simply serial/parallel- rather
talk of efficiency
Hybrid models of stimulus
processing in visual search
