EPS London (2012)
-
Upload
jimgrange -
Category
Technology
-
view
500 -
download
0
Transcript of EPS London (2012)
The Effect of Task-Relevant andIrrelevant Verbalisations
on Inhibition in Task Switching
James A. Grange & Justine Carine
Cognitive Inhibition
“...the stopping or overriding of a mental process, in whole or in part, with or without intention.”– MacLeod (2007)
Houghton et al. (2009)
Inhibition in Task Switching
A B A
Time
Mayr & Keele (2000)
Inhibition in Task Switching
A B A
Time
Mayr & Keele (2000)
Inhibition in Task Switching
A B A
Time
Mayr & Keele (2000)
Inhibition in Task Switching
A B A
Time
Mayr & Keele (2000)
Inhibition in Task Switching
A B AC B A
Inhibition in Task Switching
A B AC B A
Backward Inhibition (BI) = RT(ABA) – RT(CBA)“N–2 repetition cost”
What is Inhibited?
• Mayr & Keele (2000) originally suggested the task-set as a whole is inhibited– Scatter-gun approach
• More parsimonious to suggest inhibition is selective– Thus targeting interference directly
• Several components of trial structure can generate interference– Cue, stimulus, response selection/execution
Houghton et al. (2009)
Houghton et al. (2009)
• Word cues verbally described the characteristic of the target to search for
• Participants must generate an endogenous representation of target
Houghton et al. (2009)
• Iconic cues visually described the characteristic of the target to search for
• Environment provides bottom-up support of target
selection as cue provides allessential information for taskperformance
Houghton et al. (2009)
• Abstract cues totally unrelated to characteristic of the target to search for
• Environment provides no support for target selection
• More processing required in WM
Houghton et al. (2009)
• Modulation of inhibition cannot be due to response- or stimulus processes– Response mappings did not change throughout
the experiment– Stimulus display did not change
Houghton et al. (2009)
Error bars denote +/- 1 Standard Error around the mean
What is Inhibited?
• Results suggest cue-related processes affect inhibition– Simpler cues provide exogenous support for task
preparation– Greater preparation leads to less observable
inhibition?
• Relatively underspecified area of the literature
Cumulative Distribution Functions
• Provide novel investigation of effects of preparation
• CDFs provide overview of an effect across the entire RT distribution – Rank participant’s RTs for each condition from
fastest to slowest– Calculate percentile cut-off points for each
condition separately– Average these points across participants and plot
Cumulative Distribution Functions
Cumulative Distribution Functions
Cumulative Distribution Functions
Effects of Preparation
Grange & Houghton (2011) – Psych. Bull. & Review.
Effects of Preparation
Grange & Houghton (2011) – Psych. Bull. & Review.
Purpose of Experiments
• Can task-specific preparation reduce observable inhibition?
• Two approaches:– Disrupt preparation via articulatory suppression
(Experiment 1)– Encourage preparation via task-relevant
verbalisations (Experiment 2)
Experiment 1 - Introduction
• Relationship between verbalisation and behavioural-control functions well established (Vygotsky, 1962)
• Good evidence that verbal representations required in task switching (Saeki & Saito, 2004a, 2004b, in press)
“SHADED”
Experiment 1 – Introduction
• Articulatory suppression (AS) disrupts sub-vocal task preparation– If task preparation involves forming verbal
representations, AS should interfere with this• Control condition of foot-tapping – Secondary task that involves no verbal component
• Prediction: greater n-2 repetition cost with AS
Experiment 1 - Method• Used abstract cues of Houghton
et al. (2009)– N = 16– Fully repeated measures– 326 trials per verbalisation
condition
• Articulatory Suppression condition required overt verbalisation of “Blah” at 2Hz.
• Foot Tapping condition required tap of dominant foot at 2Hz.
Experiment 1 - Results
• Interaction not significant (p>.58)
• Bayesian t-test shows the null is 4.56 times more likely than the experimental hypothesis
Error bars denote +/- 1 SE around the mean
37ms
23ms
Experiment 1 - Results
Articulatory Suppression Foot Tapping
Experiment 1 - Discussion
• Suppression slowed RTs– Demonstrates AS is having an adverse effect
• No influence of AS on observable cost of inhibition– Interfering with verbal representations does not
affect inhibition• Data quite noisy– Quite slow overall RT (~800-900ms)– large variance
Experiment 2 - Introduction
• Instead of disrupting verbal representations, Experiment 2 sought to encourage them
• Two conditions: Relevant verbalisations & irrelevant verbalisations
Experiment 2 - Introduction
• Instead of disrupting verbal representations, E2 sought to encourage them
• Two conditions: Relevant verbalisations & irrelevant verbalisations
“Square” “Triangle” “Octagon”
Experiment 2 - Introduction
• Instead of disrupting verbal representations, E2 sought to encourage them
• Two conditions: Relevant verbalisations & irrelevant verbalisations
“Shaded” “Bordered” “Angled”
Experiment 2 - Method
– N = 29– Fully repeated measures– 326 trials per verbalisation condition
“Bordered” “Shaded” “Angled”
1,000ms
Experiment 2 - Results
• Interaction not significant (p>.57)
• Bayesian t-test shows the null is 5.98 times more likely than the experimental hypothesis
Error bars denote +/- 1 SE around the mean
20ms
13ms
Irrelevant Verbalisations Relevant Verbalisations
Experiment 2 - Results
Irrelevant Verbalisations Relevant verbalisations
Experiment 2 - Discussion
• Relevant verbalisations speeded performance– Encouraging relevant verbalisation aids
performance• No significant influence on observable
inhibition– Either mean RT or using CDF analysis
• Both experiments show trend in the predicted direction
Experiment 2 - Discussion
37ms
23ms
20ms
13ms
Irrelevant Verbalisations Relevant Verbalisations
General Discussion• A quandary...– Inhibition absent from fast end (prepared?) of RT
distribution– Task relevant verbalisations speed RT– Articulatory Suppression slows RT– But neither affect observable inhibition
• Perhaps absence of inhibition at faster RTs not due to preparation?– Models of choice RT may suggest otherwise...– Thanks to Darryl Schneider (personal communication)
for highlighting this possibility!
General DiscussionRatcliff Diffusion Model
General DiscussionV (Drift Rate) a (Response Caution)
Ter (Non-decision time)
General Discussion
Simulated Data
What Affects Drift Rate?
• Task difficulty• Stimulus quality• Familiarity• Stimulus frequency...
• Advanced preparation of task performance?
Thank you. Any Questions (or
suggestions!)?
www.cognitivecontrol.co.uk