Workshop 1 - Techniques (Brain scans)
The cocktail party effect
● Ability to follow one conversation in the presence of many
● Attention filters out background noise
○ selective attention
● The study was exploring where the filtering of info happens
Where is the attention bottleneck?
Early vs late selection model -
● Broadbent’s model of selective attention
● Attention starts to filter during the early stages VS everything being
processed and later filtered
● Early selection - cochlea, brainstem, midbrain (10ms and early in
the pathway)
● Late selection - Cortex
Solutions?
Does attention affect early processing (early selection), or are early parts of processing ”automatic”,
independent of attention (late selection)?
What is ”early”?
1. Early (10 ms)
2. Early in the auditory pathway: brainstem (which comes before the cortex)
How do you test whether something is affected by attention?
1. Attend
2. Unattend or Passive
Scanning techniques:
● fMRI
○ Not suitable
○ See an effect - but you aren't sure when the stimulus occurred
○ Not sure when the effect takes place - lag/ sluggish
○ Results might show that early stages of the sensory pathway are affected by attention,
although they would not be able to show precisely when this takes place with the stimulus
onset.
● TMS and tDCS
○ Not a good idea
○ Doesn't reach the deep brain
○ Can't see the midbrain
● MEG
○ No
○ Only see cortical activity
○ Can’t measure midbrain
● EEG
○ Good idea
○ Good temporal resolution and captures activity in all parts of the brain (including the
brainstem)
○ First 10ms of activity = auditory brainstem activity
○ Auditory brainstem response is affected by attention -> good news for early selection
● There is a bottleneck but it's not fixed - it alters due to the tasks