LTM Flashcards
What are the limitations of long-term memory (LTM)?
We don’t know duration of LTM
We don’t know capacity of LTM
What kinds of details do the STM and LTM retain?
Specific details of sensory stimulus
Abstracted semantic info without specific physical details
What two approaches have researchers taken to memories theories?
Systems approach: Each system does a different thing
Processing approach: Info processed in diff ways
Serial position curve
Primacy effect
Recency effect
Most items remembered at end and then at start, middle recalled last
Primacy: Every time you rehearse, you repeat beginning of sequence (allows encoding to LTM)
Recency: End items still active in STM
Memory systems framework (Squire) splits LTM into what two memories?
Nondeclarative (implicit) - Can’t talk about
Declarative (explicit) - Can talk about
Declarative (explicit) memory is split into which two memories?
Semantic - Facts
Episodic - Events
Which part of the brain is most involved in declarative (explicit) memory?
Medial temporal lobe diencephalon
What are the declarative (explicit) memory tasks?
Recognition (like MC)
Recall (like SA)
- Free recall: Report items in any order
- Serial recall: Repeat items in order presented
- Cued recall: Uses a hint to help recall
What are the nondeclarative (implicit) memory tasks?
Procedural (Perform task repeatedly)
Priming (Behav affected by previous exp)
- Repetition priming: Same stimulus presented in multiple occasions
- Semantic priming: Semantically similar stimulus presented in multiple occasions
Retrograde amnesia
Anterograde amnesia
Memory loss for events prior to trauma
- Not permanent; memory will gradually come back
Memory loss for events after trauma
- Little recovery
Patient HM
Removed most of hippocampus (medial temporal lobe) because of untreatable seizures
Had anterograde amnesia (loss of ability to form new long term memories, STM was intact)
What learning did Patient HM show?
Implicit learning (could learn how to get around new facility but couldn’t state his address)
Could be primed w/ word-completion tasks but couldn’t remember being primed
Memory for action (mirror training) improved normally but he couldn’t remember practicing
Synaptic consolidation
Systems consolidation
Reconsolidation
Happens between 2 neurons
- Takes few seconds to 2 mins
Involves multiple brain structures
- Takes decades
When memory is reactivated but over a much shorter time course
Maintenance rehearsal
Elaborative rehearsal
Repeating info to encode into LTM
Elaborating on meaning of info to encode into LTM
How does long-term potentiation help w/ synaptic consolidation?
Increases sensitivity of post-synaptic neurons by causing structural changes
Standard model of consolidation and retrieval
Consolidation is hippocampal dependent
Retrieval is hippocampal independent
Hippocampus binds info across diff cortical areas
- Cortical connections strengthened over time w/out hippocampus’ help
Multiple trace theory (MTT)
Like standard model:
Consolidation hippocampal dependent
Retrieval hippocampal independent
**But for semantic memories only
(Episodic retrieval hippocampal dependent)
What happens to episodic memories over time when hippocampus isn’t involved in storage or retrieval?
Episodic memories become semantic
Craik and Lockhart
Levels of processing theory
Depth of meaning during processing determines how likely an item is to be recalled
Categorized questions in:
Case
rhyming
sentence completion (had most correct recall)
2 types of rehearsal
(Levels of processing theory)
Maintenance: Keeps info active in STM
Elaborative: Links info from STM w/ that already in LTM (meaning-based rehearsal)
What is the problem of the levels of processing theory?
Lacks definition of deep processing
Elaboration
Distinctiveness (Isolation effect / von Restorff effect)
Creates meaningful links to STM
Diff from all other memories in LTM
- One cue tells us exactly where to find info
How is forgetting from LTM caused?
Retrieval failures (info available but not accessible)
Feeling of knowing (familiarity)
Tip-of-the-tongue phenomenon
Curd vs uncued recall (Tulving and Pearlstone)
- Ppl could remember more words once categories of words were shown
Tulving and Thomson
Encoding specificity theory
Had presentation of words (phase 1) vs free association task (phase 2)
- Expected cued recall task w/ phase 1 to have poorer results bcuz less engagement
- But it was actually better than recognition task w/ phase 2
Found encoding from STM to LTM was also important factor in memory retrieval
Transfer appropriate processing
(LTM)
Info is remembered long term when the form (visual, auditory, etc) matches how it was in encoding and retrieval
Context dependent learning
State dependent memory
(LTM)
Memory benefits when environment match between encoding and retrieval
Memory benefits when internal conditions (mood) match between encoding and retrieval
How can you improve LTM?
Deep processing (Add meaning to meaningless lists)
Organize (into categories/hierarchies)
Make info personally relevant
Generate info yourself
Use imagery (Paivo’s dual-code theory)
Use interactive images (combining all items into one image)
Spacing effect
Testing effect
Info repeated is spaced out into intervals, leading to LTM benefits
Benefit in LTM when ppl retrieve info on their own rather than passively
Transfer learning
(ANN)
For training neural networks
Weights of an ANN trained on one task are reused in a diff network to learn a diff task
