memory

1) 3 unitary stores: sensory register, STM & LTM. info flows in linear way through 3 units.
2) two main stores in the sensory register are echoic (sound/auditory info) and iconic (visual info).
3) info in sensory lasts very briefly, but it has large capacity.
4) if we pay attention to info, it goes to STM (limited capacity of 7+/-2 items & duration of 30 secs). if rehearsed, it stays in STM & info is usually encoded acoustically.
5) practicing maintenance rehearsal for long enough means info passes to LTM (unlimited capacity & can last up to lifetime).
6) when we want to recall LTM info, its transferred to STM by retrieval.

1) Supporting evidence: GLANZER & CUNITZ found participants immediately recalling 1 syllable word lists could only remember beginning & end of list, showing beginning words are rehearsed into LTM & end words are still fresh in STM.
2) Info can transfer into LTM without rehearsal: CRAIG & LOCKHART gave participants list of nouns & asked questions about the which involved either deep or shallow processing. Deep processing group remembered more, showing how memories are remembered from deep processing instead of maintenance rehearsal.
3) STM & LTM shouldn't be considered unitary store: CLIVE WEARING (damaged hippocampus) had little long term memory of events, but could remember skills, showing LTM stores & processes episodic & procedural memories differently, which resulted to tripartite approach to explaining LTM.

CAPACITY = large, info unprocessed & ever-changing format.
DURATION = iconic: 1/2 a sec, echoic: 3 secs, iconic: 250 milliseconds.
ENCODING = very little encoding & stays in raw form: iconic register - visual info, echoic register - auditory info, haptic register - touch
SUPPORTING RESEARCH : SPERLING
when participants were asked to recall only 1 row instead of 3 (after being shown letters for 50m/s), recall was much better than trying to recall all the rows.
Conclusions:
Capacity of SR is bigger than 4/5 items as they didn't know what line they would recall, but still did successfully, so all letters were available in SR.
Duration of SR is very limited because time taken to recall 1 row is quicker than whole grid, showing issue of letters fading was avoided, so better recall.

CAPACITY = approx. 7 items
DURATION = 15-30 secs (can be longer w/ rehearsal)
SUPPORTING RESEARCH : PETERSON & PETERSON
participants asked to recall trigrams & the longer the time delay, the more forgetting e.g. after 3 secs, 90% recall, after 18 secs, 5% recall, this shows that without rehearsal, duration is very short.
CODING : acoustic due to verbal rehearsal, however, visual & semantic coding can occur
SUPPORTING RESEARCH : BADDELEY
participants asked to recall either acoustically similar/dissimilar words immediately after hearing them & 60-80% could recall dissimilar words, which shows STM relies on acoustic coding.

CAPACITY = unlimited
SUPPORTING RESEARCH : participants presented images of scenes for 3 secs from general categories followed by recognition test & long term memory was very good.
DURATION = up to a life time, memories based on skills, not facts, are remembered better.
SUPPORTING RESEARCH : BAHRICK
ex high - school students asked to recall classmates w/ face & name recognition which had 90% accuracy which decreased to 80% for names & 40% for faces which shows recognition is better than recall & although memories may weaken, they can last lifetime.

1)CENTRAL EXECUTIVE = pays direct attention to particular tasks & decides how brains resources are allocated to tasks, so monitors incoming info from senses/LTM & directs it to other slave systems for processing/temporarily storing. very small capacity so can only pay attention to certain amount & cant store any data.
2)PHONOLOGICAL LOOP = stores speech-based sounds for brief periods & has limited capacity. composed of the phonological store which allows acoustically-coded sounds to be briefly stored (inner ear) & the articulatory control process which allows for verbal repetitions of items from P.S (inner voice).
3)VISUO-SPATIAL SKETCHPAD = independent from phonological loop, limited capacity & its job is to process/store images in relation to what they look like & their place in the visual field. used when planning a spatial task e.g. reaching for coffee. (inner eye)
4)EPISODIC BUFFER = limited capacity & its job is to integrate info from phonological loop & visuo-spatial sketchpad & offers sense of time-sequencing. it can record events/episodes happening & send info to LTM.

EVALUATION:
-supporting evidence: Baddely found ppl found it harder to perform 2 visual tasks at once (they had to compete for same slave system) than 1 visual & 1 verbal task (slave systems separate, so not in competition), showing they're separate & also limited capacity of VSSP.
-lack of clarity of central executive: oversimplified saying its role is 'attention' as case study of EVR (cerebral brain tumour removed) showed good performance of tests of reasoning (C.E is intact) but poor for decision-making (C.E isnt intact), which suggests there's separate components of the C.E.
-practical applications: rather than just explaining the structure of memory (MSMM) it explains how it functions, leading to greater understanding in cognitive dysfunction e.g. those w/ ADHD are given simple & brief instructions to help focus on tasks due to limited capacity of phonological loop.

1) EPISODIC
-Tulving suggested it's like an autobiographical record of our memories/experiences, the storage of episodes of whats happened to us throughout our life.
-What, Where, When?
-isnt reproductive of what exactly happened, rather constructed, so prone to errors/illusions.
2) SEMANTIC
-structured record of facts/general knowledge shared with others of external world.
-independent from personal experience, time/place of encoding (simple knowledge), but may have once had a personal context e.g. types of food or capital cities.
-big amount of info available & is constantly being added to.
3) PROCEDURAL
-unlike other 2, this is unconscious memory of skills of how to do things.
-particularly for how to use objects/physical movements e.g. riding a bike.
-acquired through repetition/practice & composed of automatic sensorimotor behaviours which are deeply embedded to the extent we arent aware of them.
-'body memories' e.g. with practice, driving a car becomes 'second nature'

EVALUATION:
-supporting biological evidence: Tulving injected 6 ppl w/ radioactive particles to track the blood flow when recalling episodic (childhood) & semantic (historical knowledge) memories & found 3/6 ppl had increased blood flow at back of brain for semantic & same at front of brain for episodic, showing different areas used for different natures of LTM.
-supporting case study: Clive Wearing (damaged hippocampus) was unable to store any new episodic/semantic memories, but could remember how to use objects/physical movements (procedural memory still intact). also found in those w/ Huntingdon's disease that they could learn new knowledge but couldnt learn new motor skills. shows patients w/ deficits in some parts of LTM still have others in tact, so must be separate.
-challenging research: 3 types should be considered 'declarative memory' as research has shown interdependence between episodic & semantic. no matter how many types of LTM there are, may not be helpful in helping those w/ cognitive dysfunctions recover. e.g. those w/ stokes disease.

PROACTIVE INTERFERENCE = old information interferes w/ new information e.g. calling new partner by old partners name. Jacoby: this occurs due to strength of old learning is competing w/ weakness of new learning.
RETROACTIVE INTERFERENCE = new information interferes w/ old information e.g. learning new car number plate means you cant remember your old one.

CONTEXT DEPENDENT FORGETTING = external cues are different from when we encoded info to when we're retrieving it e.g. learning info at home & recalling it in exam hall
STATE-DEPENDENT FORGETTING = internal states of awareness are different from when we encoded info to when we're retrieving it e.g. learning things whilst drunk & trying to recall the next day.
TULVINGS ENCODING SPECIFICITY PRINCIPLE = cues can help us remember things if they're present at times of encoding & retrieval, if they're different, forgetting is likely to occur.

EVALUATION:
-supporting research: GODDEN & BADDELEY studied deep water divers recalling 36 learnt words in 4 conditions & found following accuracy: learnt & recall on land (37%), learnt underwater & recall on land (23%), learnt on land & recall underwater (24%), learnt & recall underwater (32%) which shows importance of the same context when encoding + retrieving info.
-supporting research: CARTER & CASSADAY gave participants either anti-histamines or placebos & asked them to recall info in four different conditions. found decrease in accuracy of recall when internal states didnt match to encoding.
-lacks ecological validity: doesnt reflect real-life as our states/environments would have to be significantly different for forgetting, but only subtle changes in our environment/internal state are the norm, so we cant generalise findings.
-applications: in novel situations where correct instructions need to b vitally remembered, we can apply this knowledge e.g. soldiers less likely to remember if issued back at the safety of the camp. this means the use of different scenarios (informed by this theory) can help ppl save lives.

MISLEADING INFORMATION = our memories of events arent precise, instead composing of accurate fragments, whilst we fill in the rest of the gaps. this reconstruction of events can be further compounded by info we receive after the event, distorting memory e.g. subtle word change in a question asked.
LOFTUS & PALMER:
-experiment 1:participants viewed films of automobile accidents & then split into 5 conditions, each asking the same question w/ a different verb.
-experiment 2:split into 3 groups asking same question either w/ 'smashed', 'hit', or nothing (control). retest one week later found those who're asked question w/ 'smashed' more likely to answer yes to; 'did you see any broken glass?', even though there was none present in the film.

EVALUATION:
-ecological validity: doesnt reflect real life as witnesses of a real car-crash would produce a far greater emotional response, also those from 'smashed' group answering yes may have done so due to demand characteristics.
-reliability: other repeated experiments found similar results, however, this isnt the case when the misleading info is obviously incorrect. research where a slide of a large red purse was shown to participants who later read an incident account by an alleged 'psychology professor' who claimed it to be brown. it was found even after this, only 2 switched their answers to brown, showing the testimony is unlikely to change answer when the misleading info is clearly wrong.
-applications: informed police to not use misleading questions on witnesses & has helped development of the cognitive interview. this is economically beneficial as police can conserve limited resources by reducing likelihood of obtaining false info off witnesses.

POST-EVENTS DISCUSSION = errors in eyewitness accounts can occur after the incident when they come in contact w/ misinformation, such as discussion w/ co-witnesses.
GABBERT ET AL:
-participants watched video of girl stealing money from wallet & tested individually (control) or in pairs (co-witness, where only 1 was shown perspective of girl actually stealing the money).
-co-witness group had post-event discussion & then all filled out questionnaire.
-found 71% co-witness group recalled info they never saw & 60% said girl was guilty, despite not seeing her steal.

EVALUATION:
-mundane realism: doesnt reflect a real-life post-events discussion, as artificiality of lab condition could produce conformity affect amongst participants answers, furthermore, they were aware it wouldnt have real-life consequence (girl sent to prison) so more likely to recall conform, thus lowering validity.
-reliability: other research has found misinformation in post-event discussion, particularly delivered socially, were less likely to recall accurate info compared to control group.
-applications: police can advise for no post-discussion w/ co-witnesses & strive to interview them ASAP, they can also take into account the likelihood of post-event discussion & reconstruct their questions accordingly to gather most accurate info.

ANXIETY = when eyewitnesses become anxious/stressed (especially if victim) due to danger they perceive themselves to be in. there are individual differences in lvl of anxiety, but it can distort EWT.
-The Yerkes-Dodson effect: where anxious physiological arousal (pounding heart, rapid/shallow breathing) is too high or too low, it can negatively affect performance/recall, but when its at a moderate lvl, its beneficial for performance. it can explain mistakes in EWT for violent crimes, which would cause high anxiety.
LOFTUS: WEAPONS FOCUS EFFECT
-he believed anxiety of seeing a weapon causes all your attention to be focussed to its central detail, so less likely to recall peripheral details e.g. what offender looked like.
-used independent groups, either hearing argument in next room followed by a person coming out covered in blood & carrying letter opener (C1), or hearing conversation in next room followed by a person coming out with a pen & grease in their hands (C2).
-found only 33% could identify culprit from C1, compared to 49% from C2.

EVALUATION:
-internal validity: weapon focus effect may be caused by surprise, not anxiety. study where a video of hair salon w/ items like scissors/wallet (low unusualness) & raw chicken/gun (high unusualness) was shown found EW accuracy was much poorer w/ high unusualness, showing this could be the causation, not anxiety.
-reliability/artificiality: research into real-life crime (shop-keeper shooting thief dead) used interviews (which were the same used by police immediately after the event) on witnesses 5 months later & found those reporting high anxiety had highest lvl of accuracy (88%) compared to those reporting low anxiety, who had lower accuracy (75%), showing inconsistencies between lab experiment results & real-life eyewitnesses.
-applications: increased understanding of role of anxiety in EWT accuracy has led to police awareness, so they seek to compound its findings w/ hard evidence e.g. DNA/CCTV & dont solely rely on EWT.

1)RECREATE ORIGINAL CONTEXT: picture circumstances surrounding the crime & visualise the scene e.g. who's present/nearby etc.
Standard interview = asks to freely recall events.
2)REPORT EVERY DETAIL: tell them how ppl leave details out they deem irrelevant, so should tell whole story w/ nothing left out.
Standard interview = they repeat questions needed for EWT, thus interrupting witness.
3)RECALL EVENT IN REVERSE: tell them it may feel strange, but recall from end-beginning, or significant part-beginning, which ensures they dont recreate aspects of event in relation to stereotypes of how they perceive a crime to occur.
Standard interview = free recall means chronological order is more likely used.
4)RECALL FROM DIFFERENT PERSPECTIVE: different physical perspective to that of the victim to disrupt personal expectations of how a crime happens.
Standard interview = focusses on witness POV on what they saw.

1) memory is complex & made up from different types of memory.
2) retrieval cues are important, in their absence mistakes are likely.
3) misleading info can lead to inaccurate recall.

EVALUATION:

-supporting research: meta-analysis of 53 studies using ECI (Enhanced Cognitive Interview) found 34% avg. increase in accuracy of recall compared to standard interview, however, studies were conducted in lab settings, so possibly low in ecological validity/generalisability.

-reliability: study where students were shown police training videos & later individually interviewed by Law enforcement officers who're either using CI or standard. found CI generated significantly higher correct items & only slightly higher incorrect/made up items than standard interview.

-economic implications: CI takes longer/more training than standard, & research has found combination of certain aspects (context reinstatement & report everything) has produced most effective reports, showing that its viable to only use some aspects of CI in order to not waste resources on less successful aspects.