Are pattern separation and completion independent processes?
An individual's ability to distinguish between highly similar memories does not predict how well they can reconstruct a whole memory from a partial cue.
Source
Pattern separation and pattern completion: Behaviorally separable processes?
What they did
The researchers tested 84 undergraduate students using memory tasks with a shared format of multi-element events. To measure pattern separation, participants viewed 60 encoding trials and completed 70 test trials where they distinguished targets from similar lures and novel foils. To measure pattern completion, participants encoded 24 events and completed 144 self-paced test trials where they reconstructed events from partial cues.
What they found
The researchers found that individual performance on pattern separation did not correlate with pattern completion, showing a non-significant correlation of .12. In contrast, general memory performance across both tasks strongly correlated at .51, demonstrating that the lack of relationship between separation and completion is specific rather than a failure of the tasks to measure general memory ability.
The limits
What it doesn't show
This study does not establish a causal relationship between the neural mechanisms of pattern separation and pattern completion, as it relies on correlational behavioral data. Additionally, the sample consisted entirely of 84 young undergraduate students, meaning the independence of these processes might not generalize to older adults or clinical populations who often show age-related memory declines. Finally, the results are limited to the specific cartoon multi-element task design and might not apply to other behavioral paradigms that define these memory processes differently.
Key terms
- Pattern separation
- The neural and behavioral process of distinguishing similar experiences by assigning them to non-overlapping neural representations.
- Pattern completion
- The process of retrieving a whole, integrated memory when presented with only a partial or degraded cue.
- Retrieval dependency
- The statistical measure of how much the retrieval of different parts of the same event is mutually contingent, indicating holistic recollection.
- Mnemonic discrimination
- The ability to behaviorally distinguish between highly similar target items and lure items in memory.
- Lure classification
- The accuracy and precision with which a participant identifies a highly similar but slightly altered item as similar rather than old.
Flashcards
Want these cards to stick?
Save the deck to NoteFren and study it with spaced repetition.
Quiz yourself
What is the computational purpose of pattern separation in episodic memory?
Common questions
Why did the researchers use multi-element events instead of single items?
Most classic tasks test memory for single, isolated items. By using multi-element events like a scene-animal-object triad in both tasks, the researchers could directly compare pattern separation and completion under the same encoding format.
What is the difference between pattern separation failure and pattern completion?
Pattern separation failure occurs when you misidentify a similar new item as an old one. Pattern completion is the successful reconstruction of a true, old memory from a cue, though some older models historically treated separation failure as a false shift towards completion.
How do we know the lack of correlation isn't just because the tasks were bad?
The researchers ran a control correlation showing that basic target recognition in one task and basic pair memory in the other were highly correlated at .51. This proves both tasks successfully tapped into a shared pool of general episodic memory ability.
What part of the brain is thought to support these two processes?
Both computations rely on the hippocampus, but they recruit different subfields: the dentate gyrus is critical for pattern separation, whereas the CA3 subfield supports the auto-associative network required for pattern completion.
More on Long-term memory