Models of memory

Seite 1

-Levels of Processing model by Craik Lockhart

-This model says the way you think about (or process) information affect how well it is remembered

-There are three different levels of processing:

-shallow processing (structured) ie think about what the word looks like eg BALLS Is the word in capitals or lower xase -> least well remembered

-intermediate processing/phonetic ie think about what the word sounds like eg muck Does this word rhyme with duck?-> mediumly well remembered

-deep processing (semantic) ie think about the meaning of the word eg dog The ___ bit a bone

Craik and Lockhart said if you use elaborate rehearsal in process the meaning of the information, it will be better remembered (as you are deeply processing it) than if you just use maintenance rehearsal (ie saying the words over and over again)

AO2: Contrast with multi-store model hich says maintenance rehearsal gets into long term memory where it will be always remembered

Evidence in support

Craik and Turling

method: Participants were shown 60 words and for each they had to answer one of three questions, either structural, auditory or semantic

Then the participants were shown a list of 80 words and had to pick the original 60

17% visual question recognised correctly

37% auditory

65% semantic

They concluded the type of processing which takes place information is encoded affects later recall

Evaluation:

Limitations: -The evidence in support of the levels of processing model lacks ecological validity because it is an artifical task we wouldn't come across in every day life: it can't reliably be generalised to explain memory in everyday life

-A problem is that there is no adequate measure of processing depth. Curator reasoning is used, depth measured by number of words recalled

-It may be that words are better remembered if you put more effort in, rather than being down to depth of processing. Maybe deep processing requires more effort than shallow processing and that's why words are remembered better

Comparison: LOP is a functional model. It defines memory by its pprocesses rather than its structure which is what multi store model focuses on

Strenghts: -Levels of Processing model gives good strategies for helping with revision eg elaborate rehearsal, put meaning onto information or make stores out of it, so it has meaning, rather than just saying it over and over again

Working memory model by Baddeley and Hitch

-model of STM only

-defines memory by its processes that go on in it (different from multistore model -defines memory by its structure, similar to Levels of Processing model, which also defines memory by what it does)

-it's your "conscious awareness"

-all components have a limited capacity (they can only hold a certain amount of information at any one time)-> if you try to do two tasks that use same component, there will be problems

central executive:

-controller- controls and co-ordinates the other subsystems

-when we give our attention to something, it goes into the central executive initially and then the Central executive allocates info into the other slave systems

-can hold all types of info- visual/sound- modality free

Phonological loop

-deals with sound/speech

-we use our inner voice just before we speak (we say the words we are goint to speak with it)

-we also use our inner voice when we read, and it helps us make sense of it

-When we rehearse something in out head, we say it with our inner voice-> hear it with our inner ear, then say it again before it completely decays

Evidence in support:

Logie:

Aim: to examine the role of working memory in performance of a computer game

Method: The game manual responses with a joystick and processing verbal elemants (listening or reading). While playing participants had to do a visuospatial distracter task or verbal memory task

Results: Visuospatial distracter task worsened performance in the perceptual motor aspects of the game performance and the verbal memory distracter task worsened performance on the verbal elements of the game

Conclusion: The visuospatial component is seperate from the sound based component of working memory. They only have a limited capacity.

Paulson: investigated if the phonological loop is made up of 2 subsystems

Method: Whilst blood flow was measured using PET scan, participant had to do 2 different word tasks:

either memorise some letters (uses acoustic primary store) or rehearse sounds (articulatory loop)

results: blood flow in the brain different each task

conclusion: 2 subsystems in phonological loop, as they are in different parts of the brain

Evaluation:

Limitations: -The Central executive is only vaguely explained in this model- Even though it is considered the most important component, little is known about it. It is diffucult to test, because it plays a part in all tasks.

-Evidence in support of it is largely based on lab experiments-> lack of ecological validity

-Strength: It is a useful model for helping provide explanations and possible treatment programmes for people with learning problems, such as dyslexia

-It treats memory as an active process, rather than a passive process- compared with the multistore model, which focuses om structure of memory rather than defining memory by what it does, like working memory model

Types of Long Term Memory by Tulving

-Episodic memory- the memory of events, either personally experienced or about which we have read or heard. These memories are usually tied to a specific time and place

-Semantic memory - this is the memory for general knowledge, a store of facts and information

-Procedural memory - the memory of skills and actions, eg driving a car or playing a musical instrument

Evaluation:

-Brain imaging studies show that encoding and retrieval of episodic memories activate different areas of the brain from those activated in encoding and retrieval of semantic memories. This would suggest that there are indeed two different systems

Theories of forgetting

Decay:

This theory explains forgetting in STM only. It's the idea that when information goes into STM, so is potentially a memory, a neural trace is made. If the information is not rehearsed, the trace fades through time.

If the information is rehearsed, the neural trace occurs again until there are structural changes to the brain and at that point the memory is there forever (in LTM) (consistent with multi store model)

(note: decay can occur too in LTM if info hasn't been rehearsed a long time)

Evidence in support of decay theory:

-Brown and Peterson Trigrams- shows information decays from STM, if you can't rehearse it

Evaluation:

-It is difficult to test decay, because it is maybe its the events between learning and recalling that push out the memory and not just time

-According to this theory, once trace has gone it can never come back. However, some studies show information can sometimes be recalled later on, even though it was not initially available to them

Motivated Forgetting

-Unconscious defence mechanism (stops us from consciously feeling anxiety)

-ego pushes painful memory into our uncoscious, so we can no longer think about it

Evidence in support:

-Williams interviewed 100 women who were abused

-38% did not mention their abuse

conclusion: They had repressed their memory of their abuse

limitation: They might have been embarassed which why they did not report it

evaluation: -It is difficult to get emotional support for repression, because it occurs inside your head and has to be inferred

-It is unethical to give participants a task which makes them very anxious, so therefore repression is difficult to investigate in a lab

Displacement theory

-Forgetting in STM only (not LTM-because it has unlimited capacity)

-This theory relies on the fact that STM has a fixed capacity of around 7 items

-When the capacity of STM is full, new information pushes out or "displaces" old information

Evidence in support:

Murdock free recall experiment

-items in the middle have been displaced

evaluation: Displacement appears to be an adequate explanation for forgetting in the STM component of the multistore model

-The explanation of forgetting is too simplistic for the workin memory model, which has a number of functionally different STM stores

Lack of Consolidation

-Forgetting in LTM only

-When a memory is laid down in LTM there are biochemical changes that take place in the brain- This process or "consolidation" takes about one hour to occur. If trauma to the head is experienced whilst this is taking place, consolidation is disrupted and the memory is never laid down properly, so is lost.

Evidence in support: Yarnell or Lynch

-American football players. When they had been concussed during the game they were:

1. asked straight away when they come round detail of the game (could answer)

2. asked again 20 mins later (could not answer)

Could answer in 1, as info still in STM. Couldn't answer in 2, as memory was not cosolidated so not in LTM

AO": -Lack of consolidation theory also supported by the finding that mice can not show spontaneuous reactivation of things they have learnt if cosolidation is disrupted

-methodological criticism of the evidence (Yarnell and Lynch) in support of the theory:

-sample size used is small, and unpresentative of the general population:

we cannot use the Yarnell and Lynch study as reliable evidence in support of forgetting in everyday life

general point for all theories of forgetting: it is difficult to be sure lack of consolidation has taken place, because we cannot directly measure it happening and can only infer it happened

weakness: no controlled conditions

Interference

-When information to be remembered gets muddled up

-tends to occur when the two types of information are similar

1.Proactive interference = when old information affecrs your memory of new information eg you call your new boyfriend by your ex-boyfriend's name

2.Retroactive interference = when new information affects memory of old information eg you move house, rememberyour new postcode,but can't remember your old postcode

-There is empirical evidence for this theory - involves learning 2 lists of word pairs, but these lack ecological validity

Evidence in support: Baddeley

-Rugby players had to recall details of games they had played earlier in the season. Some players who had been injured early in the season were asked, so they had not played many games

Players who were injured had a better recall of the details than players who had played all games (retroactive interference)

conclusion: supports interference theory because the subsequent games interfered with recall of earlier games. goes against decay theory - as according to this, both players should recall equal amounts of details (as forgetting purely function of time)

AO2: many of the studies are lab studies, artifical task, lack ecological validity, can't generalise to forgetting in everyday life

-We are unsure of the mechanisms involved in interference and how and why it occurs

-We have to infer interference has taken place from research findings because we cannot directly observe it happening

Retrieval failure:

Forgetting in LTM only- This theory says info is in LTM but you cannot access it or retrieve it

We need appropiate cues to jug our memories out of LTM. There are different types of retrieval cues that help access memories from LTM

1. State: If you learn info and recall it in the same mood you recall it better

2.Content:-recall is better if you learn and recall it in same context (place) eg if you did your psychology exam in room 12 should do better than in hall

Evidence in support: Godden and Baddeley

-deep sea divers

-investigated importance of environmental context in memory and forgetting

-divers heard a list of 40 unrelated words on a beach in dry conditions or under 10 feet of water in wet conditions

-adter hearing the words they were tested in the same environment context, or in the other environment

-Participants recalled significantly more words if they were tested in same environment as where they learnt words

-if in same environment 10% better recall compared toother environment

AO2: can be criticised because extreme conditions, can't generalise to everyday forgetting

3.Organisation

-participants recall more infi if the material is organised than if it is presented randomly

eg if participants have to learn words under categories and are given category headings when they're asked to recall the words recall will be better (category headings "cue" recall) than if participants use free recall

Evidence in support: Turing and Peterson

-investigated retrieval failure using organisation

-participants weregiven a total of 48 words to remember

-the words were organised into 12 categories with 4 words in each category and one of the 4 words was the category heading-a control group of participants were to remember as many words as they could

Results: without any cues, participants in the control condition recalled approx 40% of the words, with the category headings as cues the experimenter group recalled over 60% of the words