Recovered Memory: False Memories PDF

memory 1 false memoriese Recovered Memories Recovered Memory is “the reappearance in consciousness of memories for past events after a period during which these memories were not accessible...

memory 1 false memoriese Recovered Memories Recovered Memory is “the reappearance in consciousness of memories for past events after a period during which these memories were not accessible”. Aka you may have memories you dont know you have. An Example: Professor Cheit's recovered memories for events which occurred 24 years earlier (Freyd, 1996). It is often claimed that memories for Childhood Sexual Abuse can be recovered spontaneously or as part of psychotherapy (14 interesting examples in Shobe & Schooler, 2001). Three key requirements for calling a memory “recovered” (Schooler et al., 1997) (1) Reality of event (2) Reality of forgetting (3) Reality of recovery False Memory Syndrome is “the systematic creation of memories for events which never in fact occurred”. alot of therapeutic approaches are coming to terms with childhood trauma, but of its believed that would help their symptoms, memories may be created falsely An Example: Paul Ingram's memories for suggestions from Richard Ofshe. (Loftus & Ketcham 1994) had no memory of sexually assaulting his children, but ended up writing a 3 page confession of a fake memory suggested too him. It has been claimed that these memories are frequently created as a result of careless use of "memory work" during psychotherapy (7 out of the 14 interesting examples given in Shobe & Schooler, 2001). When you study memory by remembering a list of words, the words remembered follow a pattern known as the serial position curve. This shows that you are good at remembering the first and last items read, but poorer in the middle. This is short term memory. Roediger & McDermott (1995) Experiment 1 -False list recall Got people to remember a list of words, Correct Recall: Overall 65% probability of recalling an item that was on the list. Errors in Recall: But mean overall probability of 40% of recalling a word that was not on the list. Experiment 1: Recognition Results. Standard Scoring Technique for Recognition Experiments: In a recognition test, youre presented with an item and asked if you remember it. Wether it s a new item not on the list or an old one. (Hit Rates and False Alarm Rates can be combined to create measures of memory performance based on Signal Detection Theory, e.g. d´, A´, c) -Mean hit rate (response 3 or 4) for the items that were on the list was 86%. -Mean false alarm rate (response 3 or 4) for critical items not on the list was 84%. Critical Lures were close associates of all the items on the list that was presented. Meaning the distractor, new words are close to the ones actually on the list. They used fruedian association words. Roediger & McDermott (1995) Experiment 2 Used longer lists and more of them (16 lists, each of 15 items), and participants attempted recall for only half the lists (but attempted recognition for all 16 lists). -Errors in Recall: Mean overall probability of 55% of recalling a critical item that was not on the list. -Correct Recall results again show a classic Serial Position Curve Recognition results for experiment 2 Remember/Know Judgments terms to know - Remember items are those where participants have a vivid memory for the actual presentation of the item. -Knowing items are those where participants are sure that they were on the list but don't actually have a memory for the moment of the hearing the word. In cases where the critical lure was incorrectly recalled, incorrect recognition rates rise to 93% (with 73% given “Remember” judgments). With recall, better at remembering lists that were, but worse with words that werent. DRM: Summary of Findings - “False” memories are very easy to produce in this paradigm. The result was expected (it's based on a study previously done by Deese 1959) - hence the DRM paradigm (Deese, Roediger & McDermott). It is so reliable that it is available as a standard classroom demonstration (e.g. Pardilla-Delgado & Payne, 2017), and works with other stimuli such as pictures (Dechterenko et al., 2021). Participants cannot distinguish between false and true memories in this experiment - the phenomenology is the same. Questions about Implications Freyd & Gleaves (1996) raise three key questions about the relationship between recovered/false memories in the real world and the DRM paradigm: Are words presented in a list really events? If false memories for Childhood Sexual Abuse are not close associates of things that actually happened, how is DRM relevant? Can we generalise from artificial laboratory studies to meaningful events in the real world? “Lost in a shopping mall” - Loftus & Pickrell (1995) Participants: 24 students and their parent or older sibling. Materials: Four short stories about events from the student’s childhood provided by the parent or older sibling. Three of the stories are true, the fourth is a fabricated account of being lost for an extended period in a mall or large department store at about the age of 5. Parents/siblings confirm that no such event actually occurred. Procedure: Students first fill in a questionnaire booklet that describes what they can remember about each event. They are then interviewed by a psychologist 1 to 2 weeks later about each event and interviewed again a further 1 to 2 weeks later Seven of the 24 students accepted the false memory, and 6 of them maintained it at interview. When debriefed, 19 out of 24 participants correctly identified the false event (n.b. five of them incorrectly guessed one of the true events). Conclusions & Implications Simple laboratory experiments can demonstrate false recall and recognition. These "false memories" cannot be easily distinguished from memories for events that actually happened. Many of our own memories of childhood or distant events could be mixtures of suggestions, fabrications, family stories, or other people’s memories. Human Memory does not provide a consistent record of the events that we have experienced - Memory is a reconstruction from many sources. N.B. The demonstration that False Memories are possible, in no way proves that any or all Recovered Memories are not genuine. We will return to this debate throughout the remaining lectures. Encoding 2 3 We need to consciously encode information in order to subsequently remember it - or do we? Sleep learning? Simon & Emmons (1956) During sleep participants hear questions and answers every five minutes. Participants' EEGs are recorded throughout the night to monitor their sleep. Subsequently they are asked the questions they heard overnight Overall: Performance is above chance. Learning has occurred. However this is inly true until you examine the eeg data, yes they were lying their with their eyes closed but they were scrually awake. Around 80% correct at answering questions asked at that point. Bruce, Evans, Fenwick & Spencer (1970) Present material to sleeping subjects then awaken them immediately. No evidence for memory Learning while unconscious? The absence of learning about external events while asleep, does not imply that we can’t remember internal events such as dreams (Pace-Schott et al, 2003), or that sleep itself might not play an important role in the consolidation of memories (e.g. Hahn et al., 2006; Paller & Voss, 2004). Memory During Anaesthesia? - Levinson (1965) 10 dental surgery patients experience mock crisis during surgery: “Just a moment! I don't like the patient's colour. Much too blue. Her lips are very blue. I'm going to give a little more oxygen... There, that's better now. You can carry on with the operation.” One month later patients were hypnotised Four patients produced almost verbatim reports of the anaesthetist's comments. Four produced partial reports and only two produced no recall at all under hypnosis. Some Major Problems with Levinson (1965) Serious ethical questions No control condition Suggestibility under hypnosis Experimenter not blind to hypothesis/condition No measure of degree of anaesthesia But is the phenomenon possible? Potentially, and it raises at least two issues:  1. Anaesthesia may not be total – “cocktail” issue: Anaesthetic (hypnotic agent) Analgesic (removes pain) Muscle relaxant (allows surgery) Amnesic agent (prevents memory)? The cocktail needs to be perfect. 2. Different tests of memory may reveal different evidence for memory from anaesthesia Issues with anaesthesia Anaesthesia - rather like sleep - isn’t a perfectly all or none phenomenon. Neither is memory – for example - the comparison between explicit and implicit memory. Explicit vs. Implicit Memory -Explicit memory requires conscious recollection of prior experiences. -Implicit memory is knowledge revealed by tasks even though you dont know where they came from. (Graf & Schacter, 1985) -Typical Explicit Memory Tasks = asking you to remember Free Recall: Participant attempts to remember target information without any assistance from the experimenter. Cued Recall: Participant attempts to remember the target information in the presence of some specific cue (e.g. an associate of the word he or she is trying to remember). Recognition: Participant is presented with a stimulus and must decide whether it is one that he or she was asked to remember. -Typical Implicit Memory Tests = testing your memory without directly asking you to remember anything Word stem completion: WRO__ Word fragment completion: I_C_N_E_U_N_I_L or Degraded picture naming: People are better at all these tasks if they have recently encountered the correct/full item (priming) – but don’t need to be consciously aware of the encounter at the time of testing Implicit Memory from Anaesthesia? Iselin-Chaves et al., (2005) -Depth of anaesthesia carefully monitored using EEG bispectral index (BIS ranges from 100 (awake) to 0 (minimal brain activity) Participants listen to two lists of 20 words, each presented 25 times. One word presented each 4 seconds - 70 minutes of presentation. Words are all six letters long, and each word shares a stem (first three letters) with at least four other (french) words. e.g. CHA - chaque, chacun, chaton, chacal, chatte Implicit memory tested by Word Stem Completions Method for scoring Implicit memory: Jacoby 1991- process dissociation -In an Inclusion test, you ask participants to complete the word any way they want -In an Exclusion test, you ask participants to come up with words that you havent heard in the experiment, new en When researchers used Inclusion test, you get some conscience memories and some unconscious memories. -Conscious recollection is called = R -Unconscious automatic memory = A Jacoby believed in conscious stuff you have R + A, but in uncomscious you only have A We don’t Remember what we don’t Attend to, Memory for Common Objects is often surprisingly poor: -'Weapon Focus' (Loftus, 1979; Loftus, Loftus & Messo, 1987) Stress (during a crime) causes attention focusing such that only 'central information' is attended to. For example, the attacker's knife at the expense of encoding the face. -Experimental demonstrations (e.g. Christianson & Loftus,1991) Change in attention (measured using eye tracking) in stressful situations is relatively easy to demonstrate e.g. Chapman & Underwood (1998) though memory changes can be more subtle (Wessel et al., 2000; Underwood, Chapman, Berger & Crundall 2003). -The limitations of encoding strategies Chase & Ericsson (1981) Participant S.F. Simple digit span task with 250+ hours of training/testing Based on chunking: so 3492 = "3 minutes 49 point 2 seconds - near world record mile time” -Does this improve memory generally? No. Letter span still at 6 items. Encoding Failures: Conclusions -Very little explicit information is encoded for free! No evidence for sleep learning. No evidence for explicit recall after complete anaesthesia. But some implicit learning may be possible without effort, even under light anaesthesia. Dramatic memory “failures” are easily observed for details that people have never attended to. Attending to one type of information (e.g. central) can come at a cost to memory for other information (e.g. peripheral). Even successful encoding strategies in one domain do not automatically generalise to other domains. Effective Pragmatic encoding : Encoding Types of memoru research, 1) Pragmatic - Seeking ways to improve people’s abilities to learn and remember 2) Experimental - Documenting the existence and nature of memory phenomena with observations that are systematically collected 3) Atheoretical - Characterising memory in an intuitive and informal manner; focusing on phenomena rather than explanations 4) Theoretical - Explaining the mechanisms of memory with theories, models, or metaphors that capture part of a phenomenon Current research is mostly experimental and theoretical. However, the most useful research is often pragmatic and atheoretical! Hermann Ebbinghaus (1850-1909) (Experimental, Atheoretical) -Inspired by Fechner’s (1860) Elements of Psychophysics. Which was Scientific study of perceptions of sensory stimulation. -Ebbinghaus decided that the scientific study of memory was possible using similar principles to Fechner. Is proposal was that you had to use words that meant nothing. His fundamental unit of memory that he invented was the nonsense syllable (e.g. nuh, zof, veg) a pronouncable 3 letter word with no meaning – a partly successful attempt to circumvent the influence of meaning. he set himself the task of learning lists of those words to see what patterns emerged in his ability to learn the lists. Instead of relying on introspection, he wanted to follow Fechner and find ways of measuring precisely and scientifically. the technique he used is known as ‘Method of complete mastery’ – Where he would learn by reading them out and trying to repeat them back. Measured how long it took him to learn a list well enough that he could repeat it perfectly on two occasions. to show he had improved his memory, he measured how long it took to learn a list the first time it was encountered, and then second etc etc ‘Method of savings’ – Key measure of retention was how much less time he took to relearn a list Ebbinghaus on Forgetting Ebbinghaus’ Classic Forgetting Function based on serial spaced learning of 1,200 lists of 13 nonsense syllables. you do not forget things evenly over time Ebbinghaus on Encoding Encoding information is much faster if it has been learned before (even for nonsense syllables). a learning function tells you how many times it takes to relearn a list as a function of how much previous practice you’ve had. its clear practice enhances learning, makes encoding efficient Ebbinghaus Encoding Long Lists Encoding Difficulty Increases Disproportionately with List Length Encoding up to 7 items is almost instant. Short term memory has capacity of 7 things After 7 items, doubling the list length more than quadruples the time to learn the list. But after that the rate of increase in difficulty may reduce. No evidence for any “maximum” list length The limitations of encoding strategies -Chase & Ericsson (1981) Participant S.F. Had to learn digits again and again. 45 day blocks of practice, spent more then a year learning nonsense syllables Simple digit span task with 250+ hours of training/testing He developed a strategy known as chunking, converting the digits to something to give meaning. Based on chunking: so 3492 = "3 minutes 49 point 2 seconds - near world record mile time” Meaning that short term memory was not actaully bigger, capacity was still 7 things. Does this improve memory generally? No. Letter span still at 6 items Massed vs. Distributed Encoding Basic Effect was demonstrated by Ebbinghaus on himself (1885) Detailed formal experimental data here presented from Keppel (1964) 8 Blocks of learning either all on one day (Massed) or 2 blocks a day for 4 days ( Distributed) Massed conditions MPs & MPd show steady improvements over time, more effective. Distributed (DP) conditions show cross-day forgetting. Massed vs. Distributed Retention Testing is either part of the learning (0 days later),1 day later, or 8 days later. * To match retention intervals between Massed and Distributed conditions there are two versions of the massed learning: MPd (Massed on Day 4) MPs (Massed Day 1) Test day 5 or 12 -if you learn over a long period of time you are likely to remember it forever. -if tou cram before an exam, you are likely to forget sbout it entirely 6 months later Practical Training with Massed & Distributed Versions Massed practice achieves goal soonest, but it is inefficient: (Less learning per hour encoding) e.g. Baddeley & Longman (1978) training postmen to type postcodes: One 1 hour session of training per day (1x1) is much more efficient than two 2 hours sessions per day (2x2) Long term retention was also poorer in the 4hr group (2x2) But the 1 x 1 group were the least satisfied with the training! The message - Choose a training strategy that achieves your specific goals Why does Distributed Encoding Work? It is hard! This may also be why it is so difficult to achieve and why people don’t like it. e.g. Bjork 2014 – desirable difficulties in training More formally… -Encoding variability Multiple learning in slightly different contexts produces more available cues (e.g. Estes, 1959; Glenberg, 1979) -Deficient processing hypothesis You don’t pay so much attention to recently encountered things e.g. Cuddy & Jacoby, 1982; Callan & Schweighofer, 2010) -Study-phase retrieval With distributed learning, you tend to retrieve the previous episode of learning each time you re-learn and retrieval itself benefits memory (e.g. Xue et al., 2010; Feng et al., 2019) Encoding in Context Bower et al., (1969) - words learned 4x faster if given appropriate network of meanings. What theoretical structures exist in memory? -Context (Bransford & Johnson, 1972) Context is vital for efficient encoding. If context isn’t available at encoding, it can’t be added later. -Schemata (Bartlett, 1932; Brewer & Treyens, 1981) Have a seat in my office while I check the previous participant has finished... 35 seconds later, collect participant, take them out of the room and do a surprise memory test: What do you remember being in the room where you were sitting? Brewer & Treyens found that both Schema Expectancy and Saliency (Schema Inconsistency) predict recall, even though these are negatively correlated with one another - does this support the existence of schemas? Surely you can’t have it both ways! -Scripts (Schank, 1982, Smith & Graesser, 1981) Memory for stories following general scripts: Stories containing well-scripted sequences of events are learned. Some of the items in each story are typical of the script, others would be atypical. Script: Taking the dog to the vet. Typical action: Jack put his dog into the car. Atypical action: While waiting for the vet, Jack dropped his car keys -Frames (Friedman, 1979; Minsky, 1975) Conclusions and Practical Encoding Advice Not all research in psychology is experimental and theoretical – some is purely pragmatic Efficient encoding requires appropriate spacing of learning (see Rohrer & Pashler, 2007 for practical revision spacing – Revision Lecture) Relating new information to existing knowledge is important for improving encoding. One way of doing this in revision is creating mind maps to elaborate links (e.g. Buzan, 2010).  But for mind maps to be really effective, you will need to generate them yourself. This is because of the generation effect – memory for self-generated items is much better than items you have been presented with (see Burnett & Bodner, 2014 and Storm et al.,, 2016). It also requires Effortful Retrieval – Desirable Difficulties! storage failures On the Permanence of Memory Psychoanalysis: During analysis patients may recover memories for traumatic or unpleasant events which seemed to have been lost. Issues: False Memories. Repression. Does this only apply to some specific events. Hypnosis: Under hypnosis people may be age regressed to recall lost details of their lives, or details from crime scenes. Issues: Suggestibility under hypnosis false alarms go up dramatically , yes you may recall stuff but also false stuff Does hypnosis add anything to Interviewing? Brain Stimulation Wilder Penfield's work in the 1940s on Epileptics. Cut open the skull under anaesthetic, so they remain conscious and you can chat to them whilst usinig electrodes on brain areas. When pn the visual lobes, visual stimuli was created -Direct stimulation of the temporal lobes often results in patients spontaneously reportingmemory-like events. (e.g. Penfield & Perot, 1963) But… Penfield examined 1,132 patients (520 temporal lobe patients) and only got experiential reports from 40 of them. Of these 40 many only reported vague sounds. Only 12 patients reported things that could be identified as being past experiences (less than 3% of those studied). The events reported may be closer to dreams than to memories. Loftus & Loftus (1980) -Decay of information stored in memory Mechanisms for forgetting: -failure to encode -storage: decay, interference, retrieval Decay in STM? The Brown/Peterson Paradigm -involved remembering 3 letters AKA Encode a consonant trigram (e.g. TLW) Count down in 3s from a number (e.g. 492) Recall consonant trigram Performance depends on delay Conclusion- things dont stay in short term memory unless you rehearse them. Is it decay or interference? Keppel & Underwood (1962) demonstrate that Brown-Peterson forgetting is at least partly caused by Proactive Interference rather than decay. Retroactive Interference: New learning causes forgetting of old material. Proactive Interference: Old learning causes forgetting of new material. The fact that Brown/Peterson forgetting is due to proactive interference is demonstrated clearly by the release from PI phenomenon (Wickens 1970) – a change of category brings performance close to the levels of trial 1 again. So is there decay in STM? Possibly - contrast Baddeley (2003) with Nairne (2002) Interference from Misinformation Loftus & Palmer (1974) Participants watch a film of a car accident. One group are asked "About how fast were the cars going when they smashed into each other". A second group are asked "About how fast were the cars going when they hit each other". The first group give higher speed estimates than the second group. One week later both groups are asked whether they saw any broken glass in the film of the accident… No Yes "Smashed" Group 34 16 (32%) "Hit" Group 43 7 (14%) “The Misinformation Effect” Loftus (1979) interprets her results as showing that the original memory itself has been distorted by misleading post-event information. This is extremely important for work on eyewitness testimony and on recovered memories because it implies that false components of memories can be added by an experimenter / interrogator / therapist. Trace Destruction? Loftus & Loftus (1980) argue that eyewitness testimony results such as those reported by Loftus, Miller & Burns (1978) demonstrate that the memory trace can be irrevocably altered by subsequent information. Does the fact that memory is impaired as a result of misleading post-event information mean that the original information has been destroyed? 195 students watch a series of 30 slides depicting a car accident. Critical slide contains either a yield (give way) sign or a stop sign. Participants then answer a 20 item questionnaire including the question: “Did another car pass the red Datsun while it was stopped at the stop sign?” or “Did another car pass the red Datsun while it was stopped at the yield sign Lotus, miller e burns (1978) Results: After 20 minute filler task participants are tested on a series of 15 slide pairs (including the critical one). Where question was consistent performance was 75% Where question was misleading performance was 51% Effect is increased with delay (2 weeks vs. 20 mins). Reduced by forewarning or blatency, but... Unaffected by incentives ($25) - Loftus 1979. But: The misinformation effect never seems to work on all the participants in the misled group. Could it just be a form of response bias for participants where no initial memory was encoded. (thus no destruction of the original memory trace required)  Active forgetting Permanence of memory Standard Paired Associate Learning: 48-PART, 26-BOOK,17-HORSE 24 people, 20 pairs to learn each (480 items) Four Week Delay then Testing by Recall, Recognition & Relearning About half the pairs forgotten At (Cued) Recall 232 items are forgotten But does this mean theyre lost permanently from memory? To test this he did a recognition test, and 112 could be recognised. Of these 120 are not Recognised But when these 120 items which can't be recalled or recognised are relearned there is a substantial advantage for learning the old associates rather than new ones: New Associate Old Associate Performance 12/60 (20%) 30/60 (50%) Apparently forgotten memories can still influence behaviour Forgetting may be a progressive reduction in availability through interference (or partial decay) rather than a deletion of the memory Storage (Retention) Failures? Human memory is certainly not always perfect but examples of complete loss from storage are hard to find. Amnesia associated with Dementia, is one clear example, but not all amnesia does show permanent loss from memory e.g. Retrograde traumatic amnesia; Psychogenic amnesia (Kapur,1999) Even apparent failures of memory usually don’t provide clear evidence of complete trace destruction. Expert Mnemonists demonstrate that astonishing amounts of material can be stored, virtually forever. Normal forgetting may be more associated with a progressive loss of availability for individual memories due to interference. Does everyone forget? ‘S’ - Luria 1968. -appears to have almost unlimited memory for numbers and equations -memorised after a few minutes -perfect surprise recall 15 years later. S had. O training, relied on imagery, synaesthesia etc However, he struggled in other areas, he struggled recognising faces. The reason why he struggled to recognise peoples faces is because faces are so variable to change all the time, because of his memory, he would learn every single expression on their face separately, and then find it very hard to distinguish between them. he may have been remembering these things perfectly, but he wasn’t deciding meaning to them. When Vygotsky asked him to memorise the names of birds. He memorised them perfectly but didn’t understand they were birds names that he was memorising. This created the idea known as the paradox of the expert. -it should be the case that all of the things you’ve learnt before make it hard to learn anything new as we believe in retroactive interference. However, this isn’t the case. the paradox of the expert is the question of why don’t our brains actually fill up why don’t we find it hard to learn more and more stuff? It’s actually the case that as you learn more, you will find it easier easier to learn new stuff. The more you know the faster you can learn new things , it’s exactly the opposite of proactive in variance. Its a proactive advantage It’s all to do with the meanings learnt, as when you have structures of things you’ve learnt, you have links between existing knowledge and knowledge that is acquired. you can fit in the new knowledge into the structures, and this prevents interference. Freudian Repression? Is it responsible for forgettig? Repression: An active mechanism to prevent remembering Based on Freud's ideas: “Memories injurious to the ego are suppressed to avoid anxiety” Suggested memories injurious to the ego are suppressed to avoid anxiety. Deliberately forgetting) To test this, study done by Wilkinson & Cargill (1955) Male and Female Participants are told they are doing a personality study. As part of it they listen to story containing a dream description, and have to remember Dream is either neutral, or contains fairly obvious sexual imagery with an oedipal content. Result: Men have worse memory than women, but only for the oedipal material. Freudian Interpretation: Only men find the content stressful, because only men can have an Oedipus complex. So men repress the content. As it causes anxiety. However… McCullough et al. (1976) find that if participants are not told that the experiment is about personality – there is no effect.,showing we do have control over our memory, and choose not to report stuff if,for example, you think it’s embarrassing. e.g. No unconscious repression - results are just a self-presentational bias Idea that emotional or arousing stuff is going to be repressed Explored by Levinger & Clark (1961)  Free association task with neutral (e.g. CARROT) or potentially emotional stimulus words (e.g. ANGRY) Participant then generates a second word Then record response whilst they generate second word with, Galvanic Skin Responses (GSRs - a.k.a. EDRs or SCRs) recorded to assess physiological arousal (item responses vary from person to person) -Found that Free associates to neutral words recalled better than those to arousing words. Showing worse at remembering arousing stimuli But note... This is a test of memory for associates – not memory for the stimuli themselves. Memory for stimulus words generally is better if they are arousing (e.g. Rubin, 1986). This is also an immediate memory test – if Freudian repression existed to emotional events it should show at long delays… Parkin, Lewinsohn & Folkard (1982) repeat the Levinger & Clark experiment, but with a delayed condition added - At immediate testing memory for associates to arousing words is poorer, but after 7 days, memory for associates to arousing words is better than for neutral ones. This was interpreted as “Memory traces take time to consolidate – physiological arousal increases the time for the trace to consolidate, but may improve longer-term encoding.”. Mcgaygh,2006 Perservation theory, arousal causes you to encode things more deeply. You might have problems in the very short term, because the arousal is distracting you, but in the long term its strong and consolidated. -Practical applications Reconsolidation from Subsequent Arousal Learn Swahili-English vocabulary pairs and tested with cued recall twice. At first test of your memory, if theres successful retrieval it is followed by an arousing picture. A punishment -Vocabulary learning is enhanced by negative arousing pictures immediately after (Expt.1) or 2 secs after (Expt.2) successful retrieval. But arousal does not enhance performance while restudying items (Expt.3 Conclusion -Against frued,it has actually shown arousal makes memory better, -Arousal has important influences on memory General enhancements for arousing material compared to neutral (e.g. Rubin) Also “weapon focus” – selective attention towards threat: central/peripheral tradeoffs (e.g. Christianson & Loftus 1987) Evidence for arousal-related consolidation of long term memories (e.g. McGaugh, 2006), and arousal helping the reconsolidation of labile memories (Finn & Roediger). But… Arousal generally enhances memory for items and associates at long retention intervals - no experimental support for general repression of negative stimuli/events (though cf. Freyd’s criticisms of Roediger & McDermott). Part list cueing - slamecha (1968) Encode 3 word lists. 30 rare words. 30 common words. 30 butterfly associates. Recall with context -(15 words provided to help remember) Or in control condition- (no words provided) Contrary to what youd think, Part-list context actually makes it harder to remember the list than if no prompts had been given at all! Reasons for this: -you were planning to remember them in order, and giving part of the list ruins that strategy -having some of the words visually there interfers with the recall of the others not there Context may be critical for encoding, but not all context is helpful at retrieval. Directed forgetting (Furthering part list queuing) paired associated leaning experiment Each cue is apart of a category, `IE Fruit + Orange, or Metal + silver Then practice, as practice improves memory. - will do retrieval practice, providing the first word and 2 letters of the second word, then using that to trigger the memory of the second word. When asked to do the final test, but the second word is not what was practiced, it became harder to recall. because you practiced Fruit + Orange, when you see Fruit + B, you struggle to think of banana because of association to orange. Even tho a fruit beginning with B is common sense, because you rehearsed fruit and orange, you get better at remembering one thing by getting worse at remembering something else. This shows that memory gets worse in other areas when trying to recall something else. List method Directed Forgetting: Bjork (1970), Johnson (1994) you are given a list of word to remember, then experimenter says oh sorry wrong list forget this one, here is the correct list. Does telling you to forget a list have any affect? Yes. The list told to forget has much worse result, as you are told to forget. And with the new capacity, you learn the new list. Interpreting directed forgetting Item-Method Directed Forgetting: This yields substantial REMEMBER - FORGET differences that can be observed in both Recall and Recognition Generally interpreted in terms of selective rehearsal of TBR items - i.e. an encoding effect rather than inhibition  of items in storage. List-Method Directed Forgetting: This yields large recall deficits for TBF lists relative to TBR or control lists Result clear in recall, but often not observed in recognition tests. Generally interpeted in terms of Retrieval Inhibition. Items remain in memory (see intact recognition) but are actively inhibited from being recalled Practical applications of inhibition Success in inhibition appears to be correlated with active engagement of prefrontal cortex in suppressing hippocampal activation (Anderson & Levy, 2009). Individual differences in ability may explain variations in recovery from trauma. Subsequent studies have shown that inhibition paradigms can be extended to memories for real events - e.g. actions (Sahakyan & Foster, 2009), autobiographical memories (Barnier et al., 2007). Active suppression through NO-THINK or Directed Forgetting could potentially explain loss of memories from Childhood Sexual Abuse (e.g. Gordon & Connolly, 2010). In everyday situations inhibition may be important for successful retrieval, and other domains such as creative problems solving (e.g. Storm, 2011). retrieval types of Types of retrieval 1. Free Recall “What did you learn in the first memory lecture?” 2. Cued Recall “Roediger and ? (1995) did a false memory study” 3. Recognition “Was it Smith/Loftus/McDermott/Ceci?” 4. Confidence “How confident are you in your answer?” These are all explicit retrieval, asking you to go back and think of something Free recall is hardest, recognition is easiest -All require explicit retrieval unlike Ebbinghaus’ favourite measure: he used strictly implicit tests Relearning “It was McDermott …. Who was it?” Relearinging to unlesrn something again doesnt require explicitly retrieving informaton from your memory. This is even more sensitive (e.g. Nelson, 1978) but may not require explicit retrieval of the learning episode even though it is enhanced by it (e.g. Finn, Roediger & Rosenzweig, 2012) Why are some easier then others? Free Recall vs. Cued Recall Tulving & Psotka (1971) Lists of 24 words (4 members of each of 6 categories) e.g. Category - mammal Members - dog, cat, horse, cow Lists tested by recall with (cued) and without (free) category cues. Aka, saying mammal and then asking you to come up with words. Or just remember the whole list of 24 Then youll be given lots of lists to try and remember, and they will be mixed up. This is retroactive interference. =thing that you have learned AFTER the original list might interfere with retrieval In free recall, of the items you just looked at straight away, you can get between 16-24 items If you learn one more list, then ask you to go back to the first one, performance falls massively down to 12 items By the time the list goes up to 4, can only remember 6 or 7 items =Retroactive interference, things you learn after something interfers with ability to retrieve BUT, if i give you the category names, performance gets much better Therefore cues prevent retroactive interference. Cues are vital in understanding all retrieval from memory Strategies that provide cues are some of the most efficient for mnemonists: e.g. Method of Loci, Pegword Method Recall Vs Recognition A question raised is, after looking at the relationship between forms of recall, are they all part of the same thing? The Generate-Recognise Theory of Free Recall (Anderson & Bower, 1972) The idea that when we try to recall things, we're trying to came up with cues to generate items from memory. When we do generate them, is is something that is familiar? If so, we say yes. This theory assumes that recognition underlies all other forms of memory. But this implies that we can only recall items if we can recognise them – is this true? Maybe not… Even though recognition is generally easier than recall, there may be situations where we can recall items from memory that we don’t recognise Tulving & Thomson (1973 “Recognition Failure” - recalling stuff you dont recognise Thompson paradigm was another paired association -pairs like train- black. Ground-cold. Very weak associate cues, but easy to imagine. The phase two recognition process is weird in comparison to the other paradigms,. They give you a word like white, and ask you to generate 4 words associated tonthr word white. Because of being primed before hand, one of the likely words will be black, but they will do it without being aware of it Then you ask them, were any of the four words you just generated on the original list Then a classic recall condition Summary: Recall is better than recognition in this task Many words are recalled that were not recognised Why? Cued Recall task is pretty easy - The low semantic associates form good cues, easily imageable, straightforward task. Recognition task is very difficult - All the close semantic associates seem very familiar (similar to Roediger & McDermott, 1995) Self-generation effect enhances familiarity (Slamecka & Graf, 1978). Conclusions from Tulving & Thomson (1973): -You can generate it from the que, but you dont recognise it as what youve learned before. Recognition is hard because familiarity is there in all terms. Recall can produce better memory than recognition if it provides better retrieval cues. Sometimes the item itself isn’t actually the best cue for identifying the context in which it was previously encountered. The generate-recognise approach may often be used in free recall tasks, but it is not a complete model of all recall (e.g. cued recall). The Encoding Specificity Principle (Tulving, 1983): “Memory performance is best when the cues present at test match those that were encoded with the memory at study.” Its not about recognising something, its about understanding the context in which you enountered it– but we aren’t great at remembering context… Source Monitoring Problems Johnson, Taylor & Raye (1977) - developed the source monitoring framework to look at how good we are at remembering the source of information Paired Associate Learning Pairs of items are studied 2, 5, or 8 times Same pairs are also tested 2, 5, or 8 times Participants then have to judge how often each item studied and how often item tested -The two judgements are highly interdependent – participants can’t keep study and test contexts separate in memory. Can people keep real and imaginary contexts separated? Suggested menories: Goff &Roediger 1998, real and imagined events day1- you hear action statements. EG, break the toothpick. Toothpick is in front of you, and you are aske to imagine doing it Day2-Imagine this 1,3 or 5 times then, on the test session you are aske to go back and remember all the different actions an if you actually did them. Results: Shows Imagination inflation, if you imagine thing lots you become systematically more likely to believe you did it. Spotting suggested memories 16 Psychology students then given 21 real and 21 misinformed descriptions. -same stdy for misinformation affect, stop& yield sign Stimuli as in Loftus, Miller & Burns (1978) road accident. Misinformation from question “Did you see the Yield sign?” Then add a description condition Supports suggestion from Johnson, Foley, Suengas & Raye (1988) that real memories may have more Perceptual information than imagined ones. But not their prediction of more context – in fact, false context (cognitive & function) may be added for suggested memories. Classification 59% correct for real memories, 60% for suggested memories. i.e. Just better than chance, but 40% incorrect! Confidence was the main reason given for classifications (but often unreliable), sensory or geographic information supported real classification, while function and rationalisation supported suggested classifications. (geographic and cognitive information particularly misleading,they thought tgese must be real) Training judges improves discrimination accuracy, but… Time blurs the distinction between perceptual and contextual information (Johnson, Foley, Suengas, & Raye, 1988) and repeatedly thinking about events may additionally decrease the differences in memory between real and imagined events. (Suengas & Johnson (1988). Confidence & Accuracy Laboratory experiments rarely demonstrate convincing relationships between confidence and accuracy. e.g. Robinson & Johnson (1996). You are more likely to describe your thoughts when youre wrong! -Participants watch a film of a crime. Then either have to Recall or Recognise (4AFC) items from the film in a memory test, and rate their confidence in the judgements. Correlations low, but significant. Correlations better in recall than in recognition.  Ease of retrieval may explain confidence. This creates a problem with repeated recalls. In real life Confidence-Accuracy correlations may be much higher: Gruneberg & Sykes (1993) - though cf. Brewer & Wells (2006). Types of Retrieval Summary There are many ways of explicitly retrieving information from LTM. Relevant cues at the time of retrieval are vital for successful retrieval. Retrieval is best when cues at encoding match those at testing. Cues aren’t obvious – recall can occasionally be better than recognition. If cues weren’t encoded, they can’t be used at retrieval (source monitoring). Real and imagined events may have highly overlapping cues. It is not easy to spot that retrieved events were imagined/suggested. Me People may over-rely on a witness’ confidence in their memory retranial improving Hypermnesia - Erdelyi & Becker (1974) A Study having children learning poems, come back two days later and their retrieval was actually better. There was an improvement in memory. It also made an you are having the same person tested on the material repeatedly. He was curious on wether this was to do with the imaging nature of the poems. He looked at words and pictures and wether or not you spent the time in between thinking about the material or not The results that adelle found was that the number of pictures you retrieve each attempt you try, increases. So rather then memory getting worse over time it actually seems to increase. Ardelyi termed this Hypermnesia- the idea you might remember things actively you couldnt remember before. In some cases, typically involving repeated recall attempts, and often using visual stimuli net gains in memory can be observed over time. In some cases additional “thinking” can help. Meaning that In most cases hypermesia requires consistent increases in “retrieval effort” (though see Mulligan, 2006) He believed there are specific processes to get information out of memory, he proposed Normal memory may be the result of two processes: Forgetting + Reminiscence (Overall performance is how much youve forgotten but how much you have also remembered) In cases where forgetting is low, and reminiscence is high, a net gain in memory over time (hypermnesia) can be observed Does this have practical uses? This a study looking at the recall for crime. Watch a 2 minute tape where a burglar breaks into a house and shoots 3 innocent victims. Recall immediately (trial 1), then do questionnaire, then recall trial 2, then do second questionnaire, then recall trial 3. Limited recall time of 7 minutes Return 48 hours later for recall trial 4. Matches the findings, number of correct items recalled increases steadily with each recall attempt, even when there is a 48 hour delay. more correct memories without increasing the number of errors Issues with repeated testing Scrivner & Safer (1988) Recall does seem to improve over time, though this may be partly down to limited recall time (7 minutes) and the 47 box detail procedure. The second time you do the test, those first 20 items will be hugely strengthened as retrieval improves memory the most. Meaning the next 7 minute period can be dedicated This shows what adyle says, as retrevial prevents forgetting. Studies replicate hypermnesia effects for emotional items suggesting hypermnesia may be even stronger in negative arousing conditions (Kern, Libkumen & Otani, 2002). Could this be a basis for recovery of “forgotten” emotional memories? Yes it could, but be careful As repeated focus on details can make false (suggested) memories more likely with repeated testing (Lane et al., 2001). Theoretical interpretation of hypermnesia remains controversial, though it has also been shown in Recognition tests (e.g. Groninger & Murray, 2004) so it can’t just be down to retrieval effort. What else can we do to enhance retrieval – The Encoding Specificity Principle. Memory is best if you match cues at time of encoding and cues at time of retrieval. Evidence of this: Context dependant memory Interested in the stress levels of divers,Godden & Baddeley (1975) - Divers' Memories. A student suggested it is because they learn how to use equipment in a non stressful environment, but use it in a stressful environment and that is why the stress levels increase. Proposed this study with stress at both learning and retrieval. learn a lost of words either in or out of water  Then asked to recall either in or out ~memory is better in the environment learnt in, if learnt-on land, remembered more on land. Therefore, if you want people to remember they have to match the coding environment. State dependent memory How does the way you feel affect retrieval ? If someone drank alcohol and studied, if they had alcohol whilst taking an exam would it help retrieval? Alcohol generally impairs memory, and is not good for learning, however… study found that if studied whilst drunk, retrieval is better when also drunk. Further evidence for encoding specificity principle. Mood and memory scoring Changing facial expressions has appeared to change mood and mood also affects retrieval. Mostly positive material is remembered when positive. If you hold a particular facial expression, your mood tends to correlate. Different expressions require different muscles, can fake a smile by holding a pencil in your mouth to move that muscle for a smile. Minir,y changes mood. Possible results: -people have a positive bias in memory Generally more positive than negative – Pollyanna Principle Voltaire’s Candide Things you learnt whilst smiling are retrieved better when smiling - mood congruencey memor Mood Congruent Memory, We tend to recall information congruent with our current mood (i.e. if we’re happy, we think happy thoughts). Mood induction typically using Velten procedure. looking at people they induced a mood, either induced elation or depression. This involved reading personal statements, such has i have lots of friends or my friends hate me, and are asked to think about the statements. If you give a list of words to learn, positive statement participants recalled more positive words and vice versa. This is a very robust result than and works with both normal moods- e.g. everyday ambulatory monitoring – (Loeffler et al., 2013) and extreme mood states e.g Williams autobiographical memories in depression & dysphoria. This is mood congruency tho, not mood dependency. Mood congruent memory= matching conditions at retrieval to the item. Mood dependent memory -To demonstrate mood dependency you need a fully crossed design (cf. Godden and Baddeley, 1975) with neutral words (not emotional words) as the stimuli. -not replicable or reliable Mood isnt actually intergrated into encoding or retrieval, it is too unrelated. So placebos like using th same  perfume when studying and sitting the exam does not work. However remember sometimes the context can be important (refer to diving or drinking) The cognitive interview technique -A way to maximise the chances of accurate memory recall -Mentally reinstate the context of the event. Recall the scene, the weather, what you were thinking and feeling at the time, the preceding events, etc. 2) Report every detail you can remember even if it seems trivial and irrelevant. 3) Report the episode in several different temporal orders, moving backwards and forwards in time. 4) Try to describe the episode as it would have been seen from different viewpoints, not just your own. 3 and 4 are ways to find more context to provoke context Does this work ? -When applying this technique, a quarter more recall was made then in regular interviews. Social influences on retrieval. -Ross (1989) Attitudes Remembered Attitudes: Ross was interested in not the current attitude, but when asked to remember what your previous attitude was. He provided a video about dangers of jogging, injury, all the bad things about it etc, and then asks about attitudes towards excersise was last week before seeing the film After an attitude change manipulation, when people are asked what their previous attitude was, it tends to change to match their current attitude. ‘I never thought jogging was a good idea’ Memory of the past changes to make it congruent with your current attitudes This can also be done with behaviour Remembered Behaviour: After attitude change manipulations, people’s memories of their previous behaviour can become distorted (e.g.Frequency of toothbrushing, after “dangers of frequent toothbrushing” film) People Change their attitude towards tooth brushing when after video ask them to recall their behaviour this time. -After the video people say things like i only brush a couple times a week. Before the video people would say they brush religiously. The previous studies all showed memories of past events that were overly consistent with current behaviour. This could be a simply explained as a variant of state-congruent retrieval. Remembered Abilities (Conway & Ross, 1984) you wont just make your current memory congruent with your past memory, you will use your memory to justify all sorts of things in your life. -Groups of students rated their study skills before participating on a training program. Such programs are rarely successful, and this one was no different e.g. no improvements were observed (in skills or grades). Afterwards students attempted to remember their pre-course ratings. They systematically remembered their pre-course ratings as having been worse than they really were. this is because of cognitive dissonance, they want the course to be useful and not a waste of time “Getting what you want, by revising what you had.” - to make yourself happy you remember the past differently. Here retrieval is socially motivated to be state-incongruent. Remembering may complex social and tune. Improving retrieval Summary Memory can improve over time with repeated testing (Hypermnesia). Two combined processes – forgetting + reminiscence. Repeated testing can be valuable if done carefully. Applying the Encoding Specificity Principle. Context Dependent Memory, and State Dependent Memory Effects Practical application to the Cognitive Interview Technique. Emotional Congruency is important, but Mood Dependency Less So. General Social influences at the time of Retrieval can be critical. autobiographical memory using memory in everyday life. Remembering our lives : Cue word technique, Galton (1879) Memory for your own life is known as AutobiographicalMemory - memory for events as personally experienced. The technique is a way to regulate the way people remember things. Galton did this by give people a word, and get people to think of a relevant event in their life. a he found that if you do this and ask people also when they happened, you get a traditional Ebbinghaus Forgetting Function Most memories are fairly recent, the longer ago they are they get steadily less frequent. -However it is difficult to ensure these events are true eg false memories. Youd need a systematic record of when events happened to know -The most common way of doing this has been to do large scale studies of YOUR OWN memory Linton (1975) did this by recalling an event each day and writing it down and sticking it in the box. Shed pull a random note out and test her memory of these. Only remembering something ONCE after a long time period, the chances of forgetting it are almost 70%. - Initially you remember it, but the forgetting gets high. Pulling the item out multiple times over years, chance of forgetting went down by 15%. Shows you forget rapidly unless retrieving repeatedly, you can retreieve a memory just once a year and youll remember it for life. Wagenaar 1986 -repeating lintons study more technically Samples approximately one event per day, over a period of 4 years and records details of the event. Each event contains 4 cues (who,what,when, where and why) plus one critical detail, and is rated on three additional dimensions. 1. Salience 2. Emotional involvement 3. Pleasantness Each event recalled once only - testing takes one year! Cued Recall testing with 24 different cuing order's. Results standard forgetting function, higher rates of forgetting in first 6 months. But items were always recognised. Overall memory was best for recent, salient, emotional, and pleasant events. Other analyses of the original data suggest good memory for unpleasant self-critical events (Wagenaar, 1994) - not consistent with repression. Others find intensity of emotion is more important than valence for producing good memory (e.g. Congleton & Berntsen, 2019; Talarico et al., 2004 Better at remembering rare events. Emotion enhances memory of own life He categorised unpleasant events that look bad on him, self-critical, or just unpleasant events that arent self critical. Memories are better for self critical events, not what youd expect from a repression perspective. Evolutionary perspective, learning from mistakes. Issues with trusting this, how can we know they truly recognise these events, when all they say is yes? This doesn ttll us anything, we cannot know. This study investigates that. False recognition Barclay & Wellman (1986) They recorded events that happens everyday, but gets mates to also record events that DID NOT happen to act as plausible distractors. when it was a true event, up to a year later you correctly recognise everything. But when it comes to foil events, over time the chance of saying it did happen increases. therefore, While people remain pretty good at recognising their own diary entries as belonging to them, over time they become more likely to falsely accept altered events as their own. We become worse at distinguishing reality. Fantasy-prone individuals may actually be better in this sort of recognition task! (Horselenberg et al., 2004) Everyday memories can be much worse, -This study shows every aspect of your day with a go pro on, mostly just walking around. -If shown two different clips, can it be deciphered which was their own day? Misra et al., (2018) -one clip walking around the streets, one inside a large building -People are almost unable to distinguish videos from their own walks from videos of other people (as long as the weather conditions are similar). -theres no memory for details in mundane events. Most of our lives never make it to our 1 autobiography memory. Early Autobiographical memories. Most people have first memory to be from about 3/4. Studies using the cue word technique reveal surprisingly few memories from the first few years of life e.g. Waldfogel (1948). Freud referred to this as Infantile amnesia. Now known as childhood amnesia. Why dont we remember? Freud says its because of repression distressing memories like breast feeding Science today says theres no way to prove, we cannot tell if these are correct. Usher & Neisser (1993) get round this problem by using parents to verify specific events that happened in childhood. gave them health screenings about health events and tested their memory of those events. They then contacted parents to confirm events. Negative events generally well remembered. But only when they happened after age 3. Youll remember the birth of a sibling only after 3. -Problems with this study: Are the memories correct? 61% of memories in Usher & Neisser's study were confirmed by a parent. In 22% of cases the parent's memory conflicted with the child's. Are these real autobiographical memories – could they be based on family narratives or informed guesswork? AM over the lifespan, Reminisce Peak: it is evidence there is no memory for early years of life, If you ask 70 year olds to recall their whole lives, you get the reminiscence peak. This is the late childhood/early teenage years and 2os where autobiographical memories stay with you fir the rest of your life. Model of AM Conway & Pleydell-Pearce (2000) Autobiographical Memories are not simple, single Episodic Memories. a They can involve specific episodic memories (Event Specific Knowledge) but are retrieved by generating something using themes and periods within your life story. The information for memories is mixed with alot of sources. o Specific episodes can play multiple different roles in different Autobiographical Memories at different times Autobiographical Memories are transitory mental constructions within a Self-Memory System (SMS). Retrieval can be direct or generative, but all involve constructive processes (Harris & Berntsen, 2019). Retrieval is done with reference to a Working Self. A concept derived from Markus & Nurius’s (1986) theory of “possible selves”. The working self maintains our current self-concept and goals. Thus one good predictor of accuracy in dating memories is degree of self-reference (Skowronski et al., 1991) But reliance on Working Self produces the possibility of inference and bias errors (Hyman, 1999; Schacter, 2001). Autobiographical memories can change because they are generated differently when social or personal needs change The Self-Memory System (SMS) is about more than Memory. Autobiographical Memory is memory for our past life But the same SMS is intimately involved in understanding our present selves And in our understanding of our future selves. They are all involved in giving a sense of self and purpose. Its a dlexible transient thing to help solve problems.