Memory 5: Retrieval Flashcards
(13 cards)
What is retrieval?
The process of recovering a target memory from LTM is based on the activation of a retrieval cue, which subsequently activates the target memory and brings it to awareness.
Activation progression from one (or more) cue(s) to a target memory, via associative connections, and the process of spreading activation.
What are the four basic principles of retrieval?
Memory traces
Associations
Spreading activation
Retrieval cue
What is the role of memory traces in retrieval?
Vary in activation level (target strength)
Deeply encoded information, more durable memory trace, easier to access memory
Higher activation level = greater accessibility
What is the role of associations in retrieval?
Links between concepts
Vary in strength.
Stronger association = faster activation.
Vary depending on the person
What is the role of spreading activation in retrieval?
Runs in parallel to all associates.
Close semantic relation = faster activation
What is a retrieval cue?
Any aspect of the content of a memory can serve as a cue (content addressable memory)
What six things is retrieval success hypothesised to depend on?
1) Attention to cues.
2) Number of cues.
3) Relevance of cues
4) Cue-target associative strength
5) Target strength
6) Retrieval strategy/mode
Attention to cues is hypothesised to increase the activation level of a concept.
Rohrer & Pashler (2003) tested this:
2 task conditions:
- Single task: Verbal free recall of lists of 15 words
- Dual task: Verbal free recall + manual keypresses (colour identification of visual stimuli; secondary task)
More words are recalled in the single as compared to the dual task condition.
Words are recalled faster in the single as compared to the dual task condition.
This means:
Divided attention (free recall + secondary task) leads to a decrease in memory performance.
When we focus attention on just one thing, we recall this better than if attention is divided.
Lack of attention decreases the activation level given to a target memory.
Speed-accuracy trade-off (faster = more errors) not seen here as single task is faster AND more accurate
It is hypothesised that more cues facilitate better retrieval.
Rubin & Wallace (1989) tested this
2 cued recall conditions:
e.g. Learn the word cat
- Single cue:
Rhyme cue (rat) OR semantic cue (mouse)
- Dual cue:
Rhyme cue AND semantic cue
What were the results?
Dual cue increases recall a lot
Adding cues improves recall performance in a super additive fashion
One of the reasons why elaborate rehearsal is beneficial - it links the material to many different cues.
Encoding specificity principle: a cue is useful when it is present at encoding/encoded with the target memory.
Tulving & Osler (1968) tested this
- Material: Cued recall of 24 words
- 4 learning conditions: word alone (0), word (stomach) with a cue (empty) (A), word with different cue (hurt) (B), word with both cues (empty, hurt) (AB)
- 4 retrieval conditions: uncued recall (0), cued with A, B, or AB
What were the results?
Matching learning-retrieval combinations (A-A, B-B, AB-AB) lead to the best recall performance. Cue at learning and retrieval
Not the full story:
A-A and A-AB / B-B and B-AB,
are equally efficient.
AB-A and AB-B also helps, but note that AB-AB is more efficient than
AB-A/AB-B
- This means -
Cued recall performance (number of recalled words) is best when retrieval is cued
with the cue present at encoding.
A cue is relevant/useful when it is present at encoding and is encoded with the target memory trace.
It is hypothesised that a strong cue-target association facilitates activation spreading from the cue to the target.
Badre & Wagner (2007)
Task: Participants presented with two words & a cue e.g. flame/bald and candle
Which of two words is semantically related to a cue? Flame
Targets are strong vs. weakly associated.
What were the results? (Brain activations)
Increased activity in left ventrolateral prefrontal cortex (VLPFC) when the cue-target association is weak as compared to strong. (PFC is chiefly involved in cognitive control)
- Increased brain activity for weak association
- A weak cue-target association (i.e., retrieval cue activation does not automatically spread to a desired memory) is compensated for by increased activity in the prefrontal cortex, reflecting the need for cognitive control to facilitate retrieval.
- Semantic encoding through elaborate rehearsal forms stronger cue-target associations and thus facilities retrieval (our brain has to work less).
It is hypothesised that the greater the activation level of the target memory, the easier it is for a cue to activate this memory.
Wagner et al. (1998) tested this:
Task: Semantic decision about words (“abstract or concrete”).
Test: Surprise recognition (“old or new word”).
Analysis: Neural activity during encoding of words that were later remembered vs. forgotten.
What were the results?
Greater activity in the hippocampal region during encoding of words that were later remembered as compared to words that were forgotten at test. (Hippocampus is chiefly involved in memory consolidation)
- Words that were recognised correctly at test were encoded more effectively than words that were forgotten at test.
A greater target activation level (in the hippocampus) leads to better recall.
Retrieval strategy/mode may affect retrieval - it is hypothesised that retrieval is best when learning and retrieval contexts match.
Anderson & Pichert (1978) tested this:
Material: Story about two boys skipping school and hiding in one boy’s house. Detailed descriptions of the objects in this house.
Learning: Reading the story, adopting the perspective of either a burglar or a homebuyer.
Test 1: Free recall of the objects, adopting the perspective of either a burglar or homebuyer.
Test 2: As test 1, but adopting the alternative perspective.
Across participants, the percentage recalled objects is increased at Test 1 when the learning and retrieval perspectives match, as compared to when they are different.
- Within participants, the percentage recalled objects is increased at Test 2 (as compared to Test 1) when the 2nd test perspective has changed to match the learning perspective.
- Retrieval success can strategically be modulated,
e.g., is subject to cognitive/top-down factors.
- Somewhat related to the idea of transfer-appropriate processing (test performance is best when encoding and retrieval modes match).
