Can the brain pick up hidden patterns we are not aware of?
Without knowing any pattern existed, people responded faster to more predictable stimuli, and different phases of a brain response (the P3) tracked different kinds of hidden regularity.
Source
Tracking the implicit acquisition of nonadjacent transitional probabilities by ERPs
Study at a glance
- Design
- Human experiment — Within-subject ERP study: participants did a four-choice reaction time task with a hidden alternating sequence while 64-channel EEG was recorded, followed by an inclusion-exclusion awareness test.
- N
- N=32 · 32 healthy young adults; a few were excluded from individual conditions of the awareness test for not following instructions.
- Population
- Healthy Hungarian undergraduate students aged 19 to 26
- Outcome
- Reaction times and P3 peak and late-P3 amplitudes for pattern, random high-probability and random low-probability triplets; awareness test performance
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
Young adults pressed keys matching the direction of arrows across 30 blocks of trials. Unknown to them, every other arrow followed a fixed sequence, so some three-trial chunks (triplets) were far more likely than others. EEG was recorded throughout, and afterwards participants tried to reproduce or avoid the sequence to test whether their knowledge was conscious.
What they found
People were faster on sequence (pattern) trials than on random trials, and fastest of all on random trials that happened to form high-probability triplets, showing sensitivity to the hidden probabilities. The peak of the P3 was larger for pattern than random trials, reflecting the sequence, while the later part of the P3 was larger for less probable than more probable random triplets, reflecting the transitional probabilities. In the awareness test, participants still produced more high-probability triplets than chance even when told to avoid the sequence, suggesting the knowledge was largely implicit.
The limits
What it doesn't show
The reaction time differences were only a few milliseconds and did not reliably grow across practice, so the time course of learning is not clearly captured. Several P3 results ran against the authors' own predictions (for example, a larger P3 peak for predictable pattern trials), and they acknowledge that no single P3 theory explains all the effects. Performance in the awareness test differed between the inclusion and exclusion conditions, so some conscious knowledge may have developed, and the sample was small and limited to young students.
Key terms
- Implicit learning
- Acquiring knowledge of a structure or regularity without intending to and without being able to report it.
- Nonadjacent transitional probability
- The likelihood that one item predicts another that appears later in a sequence with other items in between.
- Alternating serial reaction time (ASRT) task
- A choice reaction time task in which fixed-sequence and random trials alternate, creating hidden high- and low-probability triplets.
- P3 component
- A positive ERP wave roughly 300-600 ms after a stimulus, linked to decision-making and the processing of unexpected or task-relevant events.
- Inclusion-exclusion task
- An awareness test where people first try to reproduce a sequence and then try to avoid it; failing to avoid it suggests the knowledge is implicit.
Flashcards
0 of 10 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
What were participants told about the arrow sequence?
Common questions
How do researchers know the learning was implicit?
Participants were never told about the sequence, and when asked to generate key presses that avoided it they still produced likely triplets above chance, meaning they could not consciously control that knowledge.
Why look at stimulus-locked and response-locked ERPs?
If the P3 reflects linking a stimulus to a response, its effects should appear equally when averaged to stimulus onset or to the key press; here the key effects were similar in both.
Why were random high-probability trials the fastest?
In those triplets both the middle and final items were predictable, creating a short run of predictable stimuli, which the authors suggest made them especially easy to process.
More on Learning