- Participants who had previously learned many lists of nonsense syllables (during previous experiments) were more likely to forget current lists.
-
This finding illustrates proactive interference: memory of older learned syllables interfered with the recently learned items.
Technique to investigate proactive and retroactive interference: PAIRED ASSOCIATION LEARNING TASK
- Participants learn lists of word pairs.
- Memory is subsequently tested by presenting the first word of each pair (the cue) and the participant has to give the second word (the target).
- By combining the same cue word with different target words on different learning stages (see Table 1), response competition between two similar word pairs is introduced and this may lead at test to different types of interference (or sometimes called inhibition).
- For example, when learning associations like A-B and subsequently A-D, a test A-? will result in a competition between two alternatives, B and D, whereby D is the more recent one.
-
If the task instructions specify to reproduce the target of the first word pair (i.e., retrieve B) then the second word pair (A-D) may interfere with retrieval; this is retroactive interference.
-
However, when the task is to reproduce the target of the second word pair, then the earlier formed association (A-B) may interfere with retrieval of the A-D association; this is proactive interference (also known as proactive inhibition).
Memory Interference in real life
-
Language learning- learning 2 similar languages at the same time [proactive interference]- while you are trying to find a word in a more recent language, your retrieval of the corresponding word n the native language “takes over”.
-
Skills- performing a motor skill may be interfered with by similar skills that were learned earlier and in a different context.
-
Not every skill- skills such as riding a bicycle or swimming, are very resistant to forgetting even if we don’t practice them for a long time - skills that are resistant to forgetting tend to consist of a coherent chain of actions that are performed as part of one “cycle” – so-called closed-loop actions.
-
Complex skills- first-aid skills such as cardiopulmonary resuscitation (CPR). A skill such as CPR is very complex and is applied very infrequently compared to a skill like driving a car or riding a bicycle; this may explain why the latter skills do not need refreshing while the former does. [McKenna and Glendon 1985]
Cues and context-dependent forgetting
- Two other ways in which a memory can be difficult to retrieve though it has not necessarily vanished.
-
Abernathy (1940)- (effect on memory of changes in physical environment between study and test) - students performed better on a class test when they were tested in the same classroom as the room in which they were taught.
- Godden and Baddeley (1975)
‘Unforgettable’ memories: Flashbulb memories and Self-reference effect
Flashbulb Memories
Flashbulb memory is a memory of a highly dramatic and surprising event that preserves the exact details of the context in which the observer learned about the event, as if frozen by a photograph.
According to Brown and Kulik (1977), flashbulb memories contain information about:
- the informant (the person who supplied the information about the event)
- the place where the news was heard
- the ongoing event
- individual’s own emotional state
- emotional state of others
- consequence of the event for the individual
-
they last longer and are more accurate than ordinary memories- more likely to talk about a dramatic event than an ordinary one (more rehearsed)
-
distinctness from ordinary memories
Are flashbulb memories special?
- Questioning research of Brown and Kulik
- Duration and accuracy may not be so different from mundane events
Duration
Bohannon (1988) on FBMs for the Challenger space shuttle disaster in 1986, observers’ recall of the event 2 weeks after it happened was about 80% but their recall fell to below 60% 8 months after the event. Interestingly, this reduction in recall was the same regardless of the emotional state of the participants upon hearing the news of the event: Observers who felt emotionally upset upon hearing about the event were not more likely to remember the event after 8 months than those who felt calm.
Accuracy
Neisser and Harsch (1992)- memory of challenger disaster
-
Questioned 1 day and 3 days later on 7 attributes of information
- After 3 years:
-average: only 2.95 out of 7 correct
- 25% had 0 attributes correct
- FBM was highly accurate after 3 years
This example demonstrates how inaccurate memory for an event can be despite
the observer’s high self-confidence in their accuracy of recall. It suggests that FBMs are subject to the same processes of forgetting and reconstruction as other memories. Because FBMs are triggered by highly salient, dramatic events, they are likely to be more often re-told and rehearsed than non-FBMs. These processes can cause the FBM to be supplemented with additional information that was not encoded as part of the memory in the first place. Thus, FBMs can be reconstructed just like ordinary memories. For instance, people may involuntarily add details to their memory as they recount it to others, and this act of rehearsing causes new information to be embedded in the existing memory such that the added detail becomes part of what one remembers – hence one can remember things that were not there in the first place.
In conclusion, FBMs do not necessarily differ from ordinary memories in the degree to which they can be forgotten and/or be vulnerable to reconstruction. However their apparent vividness and detail (whether real or constructed) makes them still “special” at least for the observer.
Conway et al. (1994) proposed that the personal and emotional relevance of an event is one of the defining factors that can create a FBM.
Their list of criteria to get a FBM as follows:
- prior knowledge
- personal importance
- emotional reaction / surprise
- overt rehearsal
Self-reference effect
-
Memory for information related to the self is better than information of a more impersonal kind. [This may be because we are personally more interested in self referred information, but it may also relate to the way information about the self is organised in memory.]
-
Rogers et al (1977)- had subjects make different ratings about lists of adjectives. The rating tasks differed in the level of processing required (e.g., phonemic vs. semantic) but one task also probed whether the adjective referred to the observer – thus a self-reference judgment.
-Memory for the adjectives improved dramatically after the self-reference judgments compared to the semantic or phonemic judgments, thus showing a self-reference benefit in memory.
-
Self schema- The self-schema can activate associations and new information that is voluntary related to this schema (as in a self-reference task) may be incorporated in the schema. A self-schema may act as a cue to the retrieval of the new information.
Eyewitness Testimony
- EWTs are an important part of evidence presented in the criminal justice system
- Judges and juries often have to make crucial decisions on the basis of eyewitness testimonies.
Spiro (1980) Reliability of EWT
(1) The probability of changes eyewitness memories increases with time.
(2) The confidence the subject has that their memory is accurate will be at least as high (or even higher) than for accurate memories.
(3) There is no way to distinguish inaccurate memories from accurate memories without some objective external (i.e., non-testimony based) evidence.
Post-event information
Loftus and Palmer (1974) participants were questioned after viewing a film about a car accident. The questions were prepared such that the same information was probed from the eyewitness memory but by using different suggestive words.
For example, in the question: How fast were the cars going when they hit each other ?
smashed, collided, bumped, hit,and contacted. Observers who heard the word “smashed” in the question estimated the speed of the cars at the time of the accident about 25% faster than those who heard the word “hit” or “contacted”. 1/3 of the participants who heard the word “smashed” also reported seeing broken glass in the movie of the accident when questioned a week later while there was no broken glass at all in the original movie!
How misleading post-event information distorts eyewitness reports
- Suggestive cues/questions
-
Source misattribution- this can happen when the source of the new information is similar to the source of the original information
Cognitive Interview
The cognitive interview is a technique developed by Geiselman et al. (1985) in order
to improve the reliability of traditional interviews of eyewitnesses.
-
Traditional interview: uninterrupted narrative report of the incident, guided questions for additional information.
-
Cognitive interview [maximise the amount of information that an eyewitness may recall]:
1. mental reconstruction of the environment and emotional context
2. reporting everything, including partial or incomplete information and
regardless of the eyewitness’ perceived importance of the information
3. recounting the events in a variety of orders to avoid reconstruction of what
the eyewitness thinks is likely
4. recounting the event from a different perspective
Eye-witness identification
Identity parade/line up
- Sensitive to construction and instructions
- Knowing that the suspect may be absent distorts judgements- choosing a person who most resembles the person they have in memory
-
No member of a line-up should be obvious to discard
The victim’s memory for the event can be created under highly emotional and distressing circumstances. In such cases the reliability of eyewitness identification can only be verified by reference to additional external evidence. If no such evidence can be found then the memory might be unreliable, and any court decisions based solely on unsupported eyewitness identification might be unsafe.
Improving Memory
Practice
Distributed practice- your learning of something (a skill or piece of knowledge) is broken up in separate trials that are separated by shorter or longer periods of time- [good for memory over long periods of time and more efficient]
Massed practice- learns all the information or the skill in one session without interruption, or with very brief intervals between the learning trials. [Good memory for information for only a BRIEF period of time]
-
Thus, the benefit for distributed practice over massed practice is substantial over long periods of time. This benefit is sometimes also known as the spacing effect.
The logic underlying memory aids
- Mnemotechnical methods
- Logic: meaningful associations
-If new information can be encoded in relation to some existing knowledge in memory, then an association can be formed in memory between the new and the existing information. This will improve memory for the new information because the activation of its trace in memory can be triggered by cues to the information to which it has been associated.
-
Ericsson (1988) proposed three criteria for improving memory skills
- meaningful encoding: “deep” (e.g., semantic) processing improves strength of
encoding
- retrieval structure: the information should be stored together with retrieval cues
- rehearsal / practice: practicing recall may improve the speed at which retrieval and
encoding processes
Memory aids
-
Semantic associations- remembering the colours of the rainbow
- Visual imagery
-
Method of loci- create a mental map of an environment and “stick” the items- to-remember on known locations
- Create images of the items-to-remember