Lecture 7 Flashcards
What is Rapid Serial Visual Presentation (RSVP)?
A method where items (letter, digits, words) are displayed one after the other at a rapid rate (typically 100 ms per item) in a single location under time pressure.
Processed in late selection (can process meaning even at this fast rate of presentation)
Common mistake: ‘post-target intrusions’ -> Ps report a red ‘X’ as blue, because the Y following it is blue.
Describe the attentional blink (AB) experiment
Ps are shown a series of rapidly presented single items, consisting of number distractors and two letter targets (T1 and T2). The time between the two letter targets is varied.
What is the Attentional Blink (AB)?
A temporary lapse in detecting T2 in a RSVP stream when it appears shortly after T1. Blink extended out to about T1 + 6 (200 - 600ms after 1st target).
Lag-1-sparring: Blink is not observed for T1 + 1 (100 ms is protected)
What are some additional components of the Attentional Blink? (AB)
- AB occurs if T1 and T2 are defined similarly (2 digits) or differently (red letter vs. digit)
- AB still occurs even if participants only need to detect T1, not report it.
- AB is not just a recall problem, as inference is observed even with recognition tasks.
What influences the AB?
- Surrounding items: whether or not items precede or follow T1 and T2 strongly influence AB
- Brief T2: when T2 is brief , AB is more likely
- Pattern Masks: adjacent items act as masks, curtailing processing, competing with targets for perceptual resources
In an AB task, an easier T1…
Reduces size of AB
-> a more difficult T1 increases the size of AB
What is task switching?
Task switching is the process where participants switch between two simple tasks, incurring a time and switch cost due to the executive control operation involved in disengaging from one task and preparing for another.
What are switch costs and what did Jersild (1927) study find in relation to this.
Task switch costs arise from establishing an appropriate task set and disengaging from an irrelevant set. These costs involve selecting, linking and enabling task modules while disabling irrelevant ones.
Jersild’s study showed that when participants alternated between tasks they incurred a switch cost, even when the task changes were predictable
Rogers & Monsell (1995) AABB paradigm
Task A : digit task cued by bottom boxes (press right button for odd digit, left for even)
Task B : letter task cued by top boxes (press right button for vowel, left for consonant)
Switch trials: switch from task A to B (sequence is regular and can be predicted)
Non-switch trials: task remains the same
Key findings: large switch costs incurred, where switch trials showed slower RT even though sequence was predictable
How does practice and task difficulty effect switch costs?
Practice: can reduce but not eliminate costs
Difficulty: Switching from a HARD task to an EASY task incurs greater cost (difficult to disengage from hard task)
E.g. from colour naming to word naming in the Stroop test
What factors reduce task switch costs and do they eliminate them?
Task cues and preparation time reduce task switch costs but do not eliminate them (there is always residual costs), suggesting that switch costs involve both endogenous preparation and exogenous stimulus-driven effects
What do the theories of switch costs differ in:
- The role of active preparation - endogenous factor
- The role of inference effects from prior task and task set that dissipate passively
- Whether exogenous factors play a role
What are the characteristics of automatic processes?
- Without awareness
- Without conscious deliberation (obligatory)
- Without expenditure of resources
- Fast
- Rigid/habitual
What does Kahneman’s capacity theory suggest about automaticity?
Over-learned tasks become automatic and consume fewer resources. Substantial practice improve performance, reduces task effort and facilitates task coordination
What are the key results in the Shiffrin & Schneider (1977) study in regards to automaticity
In the categorical condition (consistent mapping), participants learn targets and memory set size no longer affects performance after practice -> evidence of automaticity
In the mixed condition (varied mapping), a larger memory set always results in costs, even after extensive practice -> no automaticity