Memory

Once information reaches the memory system, it is stored in different formats depending on the memory store. Coding is the process of converting information from one form to another. 

- Baddely, lists of words to 4 groups of ps, 

1. words that are acoustically similar. eg cat, cab, can

2. words that are acoustically dissimilar eg, pit, few, cow

3. words that are semantically similar. eg, great, large, big

4. words that are semantically dissimilar. eg, good, hot, big

They read this list of words and were asked to repeat immediately, when the words were acoustically similar recall was worse. 

When asked to recall the words 20 minutes later, the recall was worst for semantically similar words. This suggests information is coded semantically into the LTM. 

A03

Artificial stimuli, no meaning to the material, cant generalise. limited application. 

The amount of information that can be held in a store.

Digit Span- Jacobs 1887, gave participants 4 digits to remember and recall in the correct order, if correct he then gave them 5, 6, 7,8,9,10 until they were wrong. This determines the individual's span. mean span was 9.3 items, letters were 7.3

Span and Chunking- Miller, noted things often come in 7 eg, the musical scale, days of the week, deadly sins etc. also noted people recall 5 words as well as 5 letters, chunking, grouping sets of digits or letters into chunks. 

A03

lacking validity, Jacobs study a long time ago, lack control eg, distraction. Confounding variables. Confirmed in other research- increases validity 

Cowan found the capacity of STM was only about 4 chunks. suggests Miller overstated, his lower end of 5 is more appropriate. 

STM- Peterson and Peterson, 24 undergraduates, 8 trials. Each trial gave 3 consonant syllables/trigrams, then given a 3 digit number to count backwards from to prevent mental rehearsal. Each trial told to stop at different times, 3,6,9,12, 15 and 18 seconds. The retention interval. Suggest STM very short unless we rehearse. 80% recall at 3 seconds, 10% recall at 18 seconds.

LTM- Bahrick et al, 392 ps from Ohio, Age 17-74. Tested on the yearbook, photo recognition and free recall. Those tested within 15 years of graduation, 90% accurate. 48 years declined to 70%. Free recall after 15 years, 60% accurate, after 48 years 30% accurate. LTM can last a long time.

A03

Meaningless stimuli, memorising constant syllables doesn't reflect real life, lacked external validity. Sometimes remember irrelevant eg, phone numbers.

Bahrick et al, higher external validity, real memories, recall lower on the meaningless material- Shepard. Confounding variables- may have looked at yearbook over the years. 

Atkinson and Shiffrin's describes how information flows through the memory system, suggests memory is made up of three stores. 

Sensory register- Stimulus from the environment, sights, sounds, smells etc, Not one store, one for each of the senses. Iconic memory (visual information, coded visually), echoic memory (sound/auditory, encoded acoustically). Duration less than half a second, high capacity eg, 100m cells in one eye storing data, pay attention it will pass further in the memory system. 

Short-term- Limited capacity store, can only concentrate on a number of things. Jacob and Miller suggest this is 7+- 2, some say more 5 than 9. Coded acoustically, lasts about 30 seconds, unless rehearsed. Maintenence rehearsal, if rehearsed long enough, passes to LTM.

Long-term- Potentially permanent store, unlimited capacity and duration. Bahrick et al, people remember names and faces for up to 50 years. Tends to be coded semantically. Once in the LTM to transfer it back to the STM we use retrieval, none are recalled directly from the LTM.

Supportive research. Baddely, mix up similar sounding words using STM but similar meaning words using LTM.  Shows different stores.

Suggests STM is one store, the study of KF, short-term memory poor when people read to him but better when he read himself. Amnesia. Other evidence for a store for non-verbal sounds. Must be more than one STM store, auditory and visual. 

MSM states more rehearsal, better remembrance. Craik and Watkins, rehearsal type is more important, maintenance rehearsal stated by MSM keeps info in STM but elaborate rehearsal necessary to get to LTM. occurs when you link new information to previously learnt information. 

Artifical material, much of the study used digits, numbers ect, 

More than one LTM store, episodic, semantic and procedural. not explained in the model. 

Tulving, multi-store model too simplistic and inflexible

Episodic memory- Store for personal events, complex, memories of events include when we met people, objects, places and behaviours involved. Time-stamped- we remember when they happened. Conscious effort to recall.

Semantic memory- Knowledge of the world, facts (where is England), words and concepts (love), not time-stamped, usually don't remember when we learnt them. Less personal, very common knowledge. Usually, don't need to be recalled deliberately. 

Procedural- Knowlege how to do things, actions/skills. don't need a conscious recall, eg driving a car. might find hard to explain if someone asks.

- HM, and clive wearing. Episodic memory impaired, semantic worked fine, procedural memory also fine. CW, play the piano. One store damaged, supports multiple.

- Tulving, PET scans, semantic and episodic located in the prefrontal cortex. Left, semantic, right episodic. Physical differences. Confirmed in later studies. The validity of findings.

- practical application, Belleville, trained old people on episodic memory, had better recall after than the control. Distinguish enables targeted treatment.

- Clinical studies, interesting, rich in detail, serious lack of control, small sample.

- Cohen and Squire said only two memories, declarative (episodic and semantic) and non-declarative (procedural)

Short term model, concerned with part if the mind active when temporarily storing and manipulating information. 

Central executive- co-ordinates the activities of the three subsystems, allocates resources to these systems. 'slave systems'.

Phonological loop- Processes auditory information and preserves the order it is presented in. Subdivided into. Phonological store- stores words. Articulatory process- allows maintenance rehearsal, approx 2 seconds of speech. 

Visuo-spatial sketchpad- Stores visual and spatial information when required. Baddeley said this has 3/4 objects. Logie divided it into subdivisions. Visual cache- stores visual data. Inner-scribe- records arrangement of objects in the visual field. 

Episodic Buffer- Added in 2000 by Baddeley, brings the material together from the other subsystems into a single memory maintaining a sense of time sequencing. A storage capacity of 4, links the WMM to LTM. 

- Shallice and Warrington's study of KF, poor STM ability for verbal but could process visual information. Difficulties with sounds but could recall letters and digits. Phonological loop damage only. Unique individual, trauma, maybe not reliable.

- Dual-task, Baddeley, ps had more difficulty doing two visual tasks than one visual and one verbal at the same time. Visual competing for the same slave system.

- central executive is seen as most important but least information about it, some believe it consists of separate components.

- Baddely, word length effect, disappears if given an articulatory suppression task. 

- Braver, brain studies, task using central executive, greater activity in the left prefrontal cortex, increased as the task became more difficult. demands on the CE increase, it is forced to work harder. 

- When two pieces of information conflict with each other, result in forgetting in one, both or distortion of memory. 

- Mainly an explanation for LTM, likely we cant get access even though they are available, interference between memories makes it harder to locate them- forgetting. 

Proactive interference- When an older memory interferes with newer ones.

Retroactive- When a new memory interferes with an old one. 

Effects of similarity- interference worse when memories are similar. McGeoch and McDonald. 

- Ps had to learn a list of 10 words until they can remember them with 100% accuracy, then they learnt a new list. 6 lists. Synonyms, antonyms, unrelated words to the original, nonsense syllabus, three-digit numbers, no list. 

- When recalling the first list, performance based on the second list. Synonyms produced the worst recall (1 word, no list 4.5), interference strongest when information is similar. 

- Lab studies, many show that both ways of interference are likely to be common ways of forgetting. High control, valid explanation.

- The artificial material, greater chance that interference will be present in a lab, lists or words (more realistic than constant syllables), not like real life. Names, faces, dates etc. 

- Baddely and Hitch, is interference better than the passing time. Rugby players asked about names of teams played against with over the season. What mattered more wasn't how long ago the games were but the number of games played since. Daily situations. Skrull also demonstrates info in daily situation, magazines, harder to recall when brands were similar. 

- Maximise interference, short time periods between learning. 

- Tulving and Psotka, 5 lists of 24 words tops, 6 categories. 70% recall of earlier words. When given the categories, 70% recall of the whole list. Cued recall. 

- Insufficient cues, when information is stored, associated cues also are, appears you have forgotten information, due to retrieval failure. Not being able to access memories there. 

Encoding specificity principle (ESP)- Tulving reviewed research, found consistent patterns in findings. If a cue is to help us recall information, it had to be present at encoding (when we learn the material) and at retrieval (when we are recalling). If cues are different at retrieval than encoding, there will be some forgetting. Other cues are also encoded at the time of learning but not in a meaningful way. context-dependent forgetting (external cues) and state-dependent forgetting (internal cues). 

Context-dependent forgetting-  Godden and Baddeley (1975), deep sea divers, 4 conditions, Learn and recall on land, learn and recall in the sea, learn on land and recall in the sea, learn in the sea and recall on land. Accurate recall 40% lower in non-matching conditions. External cues different. 

State-dependant forgetting- Carter and Cassady, anti-histamines to patients, this made them slightly drowsy (different internal physiological state from the usual alert). Then learn a list of words and recall. Learn on drugs, recall on drugs. Learn on drugs, recall not. Learn not, recall on. Learn not, recall not. When the conditions didn't match, significant lower recall. Internal cues absent. 

- Vast amount of support, Godden and Baddeley, Cassady and Carter. Eyesnck argued retrieval failure as the main reason for forgetting.

- Real life application, often environments not as vastly different as on land and in water. One room to another, likely to be less forgetting. 

- Only present when tested in a specific way, when Godden and Baddeley asked them in the 4 conditions if they recognised words, recognition the same in all 4 conditions. 

- When testing the ESP we assume words are encoded, when they cant recall, we assume this is because the word wasn't encoded, based on assumptions. 

- Real life application, fits in with our knowledge of forgetting and going into another room to remember. Police use it for witnesses at a crime scene. 

Loftus and Palmer, students watched clips of a car accident then they asked the students questions about the accident, how fast were they travelling when they __ each other? the verb was changed in each 5 groups, hit, contacted, bumped, collided, smashed. Mean for contacted was 31.8mph, for smashed it was 40.5mph. 

Response bias explanation, leading question has no effect on the actual memory but just influences the answer. Smashed encouraged a higher speed estimate. Loftus and Palmer second experiment to support the substitution explanation. those who heard smashed were more likely to report broken glass later than hit. Verb altered the memory. 

Post-event discussion- When co-witnesses discuss the crime (post-event discussion) information becomes contaminated, combine misinformation with there own memories.

Gabbert et al, ps in pairs, each watched a video of a crime but from different POV, ps could see different elements to the other. eg, the title of a book. They then discussed. 71% of ps recalled something they did not see in there video. Control group, 0%. ISI or NSI, memory conformity.

- Real life application, police interviewing, could have very serious consequences. Can improve the legal system and court.

- Artificial, films, different from real life crash, lack stress of a real accident. Yuille and Cutshall, emotions impact recall. Researchers may be too pessimistic.

- Older people less accurate than younger people, Anastasi and Rhodes, people age 18-25 and 35-45 more accurate than those 55-78 years old. All more accurate when identifying those of there own age group, own age bias. 

- Many answers those give in EWT lab studies are from demand characteristics, answer questions they know aren't right to help the researcher, to be 'helpful'. 

- Consequences of EWT in real life are not present in studies. 

Negative effect on recall

Anxiety prevents physiological arousal in the body prevents us from paying attention to important cues, recall worse. Effect of weapons: Johnson and Scott, a waiting room for a 'lab study', heard an argument in the next room, 'low anxiety' condition, the man walked through with a pen and greasy hands. 'High Anxiety' condition, hey heard the same argument but followed by the sound of breaking glass and man holding a paper knife covered in blood. The ps then picked out the man from 50 photos, 49% in the pen condition. 33% in the knife condition. Tunnel theory- narrows on the object.

Positive effects on recall

Anxiety creates physiological arousal, triggers fight or flight response. Increases alertness. Yuille and Cutshall, real-life gun shooting in Vancouver, Canada. 21 witnesses, 13 took part. Interviews 3-4 months after the incident, compared to the original police interviews. Accuracy determined by details reported, asked to say how anxious they were on a 7 point scale and if they had any emotional difficulties since eg, sleeplessness. Very accurate, little change from the original. Those who reported higher levels of stress were more accurate 88%-75%.

Explained through an optimum level, better recall when anxious until they reach a certain level, recall of event declines. 

Negative effect on recall

Anxiety prevents physiological arousal in the body prevents us from paying attention to important cues, recall worse. Effect of weapons: Johnson and Scott, a waiting room for a 'lab study', heard an argument in the next room, 'low anxiety' condition, the man walked through with a pen and greasy hands. 'High Anxiety' condition, hey heard the same argument but followed by the sound of breaking glass and man holding a paper knife covered in blood. The ps then picked out the man from 50 photos, 49% in the pen condition. 33% in the knife condition. Tunnel theory- narrows on the object.

Positive effects on recall

Anxiety creates physiological arousal, triggers fight or flight response. Increases alertness. Yuille and Cutshall, real-life gun shooting in Vancouver, Canada. 21 witnesses, 13 took part. Interviews 3-4 months after the incident, compared to the original police interviews. Accuracy determined by details reported, asked to say how anxious they were on a 7 point scale and if they had any emotional difficulties since eg, sleeplessness. Very accurate, little change from the original. Those who reported higher levels of stress were more accurate 88%-75%.

Explained through an optimum level, better recall when anxious until they reach a certain level, recall of event declines. 

- Johnson and Scotts weapon focus tests surprise not anxiety. Pickle recreated using scissors, handgun, wallet or raw chicken in a hairdressing video. Recall lower in the unusual chicken and gun. The knife is due to unusualness, not anxiety.

- Research into real-life witnesses after the event, confounding variables- post discussion etc. Extraneous variables may be responsible for the accuracy of recall. 

- Ethical issues, creating anxiety in participants, may subject to psychological harm. Questions the need for research. Could use less controlled field studies.

- Inverted U only links physical arousal to poor performance, no behavioural or emotional.

- Demand characteristics, videos of crime, most will figure out what they're asking. 

Fisher and Geiselman argued police interviewing could be improved by using better techniques when interviewing witnesses. Recommended that techniques be based on how the memory works using cognitive interviewing. 

1. Report everything, including every detail of the event, even those deemed as irrelevant. Trivial details may be important or trigger other memories.

2. Reinstate the context, related to context-dependent forgetting, witnesses should imagine themselves back at the crime scene. Imagine weather, what they could see, mood etc. 

3. Reverse the order, events recalled in a different order eg, back to front or middle to end. Prevents dishonest and expectations.

4. Change perspective, should recall events from other peoples perspective eg, perpetrator or other witnesses. Disrupt any schemas or expectations formed. 

Enhanced cognitive interview

Fisher, added additional elements, to focus on social dynamics of the interaction eg, when to establish eye contact, reducing anxiety, minimising distractions and open questions. 

- Time-consuming, more time than standard police interview. Time is needed to establish a rapport with the interviewee for them to relax. Also requires special training. 

- Milne and Bull, each technique valuable, produced more info than standard police. Combination of report everything and context reinstatement produce the best recall. Some aspects are better than others. Suggests only 2 used if not all, increases creditability.

- ECI may offer benefits, meta-analysis, 5o studies, enhanced Ci provided more correct information. REal life application, protecting society.

- Variations of Cl, police force use own interpretation of ECI.

- CL increased accurate information, recall of incorrect info may also increase. kohnken, 81% increase of correct info but also 61% increase of false info.