Memory

Atkinson and Shiffrin (1968) - First cognitive explanation of memory, previously biological means. This model explains how information flows through a series of storage systems, with 3 permanent structures in memory.

Sensory Register (SR)  - not under cognitive control, automatic response to the sensory information by the sense organs. This is the first storage system.

• Coding:
  • Iconic - Vision
  • Echoic - Sound
  • Haptic - Touch
  • Gustatory - Taste
  • Olfactory - Smell
• Crowder - SR, information in the iconic store for a few milliseconds, but 2/3 secs within the echoic store. Supporting the idea that sensory memory is coded into different sensory stores (different durations).

• Capacity
  • Large, unprocessed and highly detailed ever-changing information
  • Sperling (1960) - Investigate capacity, of sensory register
  • Flashed a 3x4 grid of letters onto a screen for 1/20th second, and asked participants to recall the the letters of one row
  • Information would fade quickly, he sounded different tones (high, medium or low), to indicate if 1st, 2nd or 3rd row had to be recalled.
  • Found - Recall of letters in the indicated row was high, which suggests all information was originally there
  • Concluded that the capacity of the SR, is large.

• Duration
  • Different amounts, decreases with age
  • Treisman (1964) - Presented identical auditory messages to both ears, with a slight delay
  • Found - 2 secs or less = noticed they were identical
  • Concluded that echoic has limited duration of 2 secs. (Iconic store - 500 milliseconds)

Temporarily stores information recieved from the SR, and is an active memory system (contains information currently being thought about). 

• Coding
  • Info raw from SR
  • Then encoded in a form STM can deal with easily
  • Research found it's main form of coding is acoustic (sound)
  • Baddeley - STM is acoustic
  • 75 participants, 1/4 word lists, x 4
  • A (acoustically similar), B (acoustically dissimilar), C (semantically similar), D (semantically dissimilar)
  • Given list containing original words in the wrong order, had to rearange to the correct order
  • List A worst recall (10%), others 60-80% recall
  • Concluding that as A was recalled worst, must be some acoustic confusion in STM, suggesting coding is acoustic. 

Capacity

• Limited, can be increased through chunking
• Jacobs (1887)
  • Serial digit span technique, P's presented with increasingly long list of numbers or letters, recall them in the right order
  • When P's fail on 50% of tasks = reached their capacity
• Results show numbers = 9 items, letters = 7 items
• Concluding the capacity of STM is limited (numbers easier to recall) 

Duration

• 30 secs, increased with rehearsal
• Peterson + Peterson (1959)
  • Nonsense trigrams read, P's had to count backwards in 3's from 300 (prevent rehearsal)
• Results show 90% recalled correctly 3 secs, only 5% 18 secs
• Concluding the duration is between 20-30 secs

Involves information over lengthy periods of time, over 30 secs, and potentially a lifetime.

Coding

• Said to be semantic
• Deeper the level of processing of stimulus, more likely it will be recalled
• Baddeley (1966)
  • 4 word lists (experiment for coding STM)
  • Given a 20 min interval before recall
  • Results, list C (semantically similar words) the worst (55%), other lists 70-85%
  • C = recalled the worst = semantic confusion, therefore coded semantically.

Capacity

• Unlimited, lost due to decay and interference
• Wagenaar (1986)
  • 2,400 events, 6 years, recall of events rather than dates
  • Results showed he had excellent recall (75% - one year, 45% - 5 years)
  • Concluding LTM is limitless

Duration

• Depends on lifespan, longer duration if originally coded well
• Bahrick (1975)
  • Investigating Very LTM
  • 400 P's between aged 17-74, set of photo's of names and faces
  • Results, those who'd left 48 years previously identified 80% of names, 70% faces. last 15 years 90% of faces and names
  • Concluding memories for faces is long lasting (VLTM)

First explanation - influential, forming other investigations into memory = +ve impact on cognitive research into memory

Research for each store and form (Jacobs, Peterson and Peterson) = Gives scientific backing and clarity

Amnesia cases - often lose either STM abilities or LTM = supporting the idea that they are seperate memory stores (fundamental point)

Over simplified - assumes that there are single stores, but WMM suggests several STM (verbal/visual), and types of LTM. = disagrees with the key ideas of MSM that there are 3 simple storage compartments.

Cohen (1990) - memory capacity can't be measured purely by the amount of info, but rather the nature of info to be recalled, as some things are easier to recall, MSM doesn't consider this = reduced validity

• Explicit - Recalled only if thought about (episodic/semantic)
• Implicit - Don't require conscious thought to be recalled (procedual)

Episodic - Specific personal event and their context

• Autobiographical record of personal experiences (when their birthday is)
• Strenth of the episodic memory, influenced by emotions present at time of coding (trauma of your child going to school for the first time)
• Highly processed events are recalled more easily, distinguish the difference between real events and imagination/ delusions.
• Prefrontal cortex is associated with intitial coding of memories
• Memories of different parts of the event are located in the different visual, auditory, olfactory areas of the brain, and connected together by the hippocampus.

Tulving (1989) - the processing of Episodic and Semantic memory tasks 

• 6 volunteers, small amount of radioactive gold, scanned to detects its location with a gamma ray detector
• 8 successive trials, each lasting 80 secs; 4 semantic + 4 episodic topics, randomised order
• Individual trials were seperated by 2 mins of rest each
• P's lie face up, on a coach with their eyes closed and indulge in either episodic and semantic memory (silent mental thoughts) 
• At a signal the P's would begin thinking about this topic, gold being injected in after 60 secs and the scanning 8 secs later, after the gold tracer arrived in the brain, scanning process took another 2.4 secs
• Findings
  • 3 P's continued, show consistent differences in cortical blood flow patterns between semantic and episodic thinking
  • Greater activation in the frontal lobes of the cortex during episodic, and greater activation in the posterior region of the cortex when retrieving semantic information
• Concluding episodic and semantic LTM involve different brain areas - biological basis

• Explicit
• All knowledge (facts, concepts, meaning) an individual has learned
• Stronger if processes more deeply, better sustained over time than episodic
• Disagreement which brain areas are used, some say hippocampus and related areas, others say there is usage in serveral brain areas

Vicari (2007) - Semantic use different brain areas than Episodic

• Case Study of CL - 8 year old suffered brain damage, removal of a tumor
• Results show CL demonstrated deficiencies in her episodic LTM functions, especially creating new episodic memories, but still create and recall semantic memories
• Concluding that episodic and semantic memory are seperate systems, using different brain areas, with the hippocampus associated with episodic memory, and the perihinal cortex with semantic LTM

• Perform learned tasks with little conscious thought (riding a bike)
• Difficult to explain in words (non-declarative)
• Occur early in life, including learning important motor skills
• Also involved in language, helping individuals to speak automatically, showing how procedural and semantic memories work together
• Associated with mainly the neocortex brain areas of primary motor cortex, cerebellum and prefrontal cortex, hippocampus doesn't need to function (like explicit)

Finke (2012) - Procedural memory exists and is seperate from episodic and semantic

• Case study of PM - 68 year old professional cellist, suffered brain damage to various brain areas after contracting encephalitis, resulting in servere amnesia
• Results show that his episodic and semantic memory LTM were very affected
• Couldn't remember musical facts (famous composers), his ability to read/ play music wasn't affected, including learning new complex pieces
• Concluding that procedural memory is seperate to episodic and semantic

Baddeley and Hitch (1974) - Questioned the existence of a single STM store, more complex than just a temporary store transferring information to LTM. STM as an active store, holding several pieces of info while being worked on

• Proposed a multi-component 'working memory' of initially 3 components (Central Executive, Phonological Loop, Visual- Spatial Sketchpad), Fourth component added in 2000 called Episodic Buffer

Central Executive (Selective Attention)

• Filter, to determine which info is received by the sense organs is and isn't attended to
• Info in all sensory forms, directs info the the model's slave systems and collects responses
• Limited capacity, only effectively cope with one strand of info at a time
• Attends to particular types of info, attaining a balance between tasks when attention needs to divided between them (talking while driving), also different inputs of info
• Baddeley (1996): generate lists of random numbers, simultaneously switching between pressing numbers and letters on a keyboard. P's found it difficult to do this at once, suggesting CE is limited in capacity and only cope with one type of info at a time

• Deals with auditory information, and the order of this information
• Limited capacity determined by the amount of info that can be spoken out loud in 2 secs
• As acoustic, confusions occur with similar sounding words

PL is divided into 2 sub parts

• Primary Acoustic Store (PAS) - known as the inner ear, dealing with words recently heard
• Articulatory Process (AP) - known as the inner voice, keeping info through sub vocal repetition

Baddeley (1975) - Capacity of PL

• Visual representations of 5 word lists for breif exposures and asked participants to write them in the same order
• 1) Single Syllable English words, 2) Pollysyllabic 
• Showed the average correct recall over several trials showed a significant increase for short words
• Suggesting that the capacity of the PL is determined by the length of time it takes to say the words, rather than by the number of items (word length effect). 

• Referred to as the inner eye
• Limited capacity, temporary memory system for holding visual and spatial information. 
• Consist of a passive visual store - Visual Cache (info in form of colour)
• Linked to the inner scribe (stores info about the relationship of items), that acts like a visual rehearsal mechanism

Klauer and Zhao (2004) - investigate existence of 2 sub systems within the VSS

• One group to carry out a visual take (Chinese ideographics)
• Other group to carry out a spatial task (memory locations of dots on a screen)
• At the same time P's were asked to carry out a spatial interference and a visual interference task, with no interference control condition
• Showing that a spatial memory task was more strongly disrupted by spatial than visual interference, and visual memory taks was more disrupted by visyal than spatial interference
• Suggesting that VSS has distinct visual and spatial components (the visual cache and the inner scribe)

• New component of the WMM, devised in 2000 (25 years later)
• The model needed a general store to operate properly, where it would store, hold, combine info not only from the visuo-spatial sketchpad, phonological loop and the central executive, but also LTM

Alkhalifa (2009) - Existence of the Episodic Buffer

• 48 uni students were presented with numbers on a screen given either as a sequence (individually) or in a parallel fashion (all at once) where participants compare info in different parts of the screen simultaneously
• Numbers of sufficient complexity were used to overwhelm the capacities of both the PL and the VSS and P's were required to answer a number of problem solving questions concerning the numbers presented
• Showing P's who were presented number in a sequence got more problem solving questions right than those in a parallel fashion. Capacity of STM was up to 25 items, exceeding PL and VSS
• Sequence was more effective, suggesting total capacity of WM is larger than capacity of PL and VSS, implying EB existence (holds items in WM until they are recalled)

Research supports it (Baddeley, Alkhalifa, Klauer and Zhao) - various components

Extremely Influential - most cognitive psychologists use the term working memory (STM) = showing it has spread throughout cognitive psychologists

Plausible model than MSM - explaining STM in both temporary storage and active processing. Also incorprates rehearsal as being on means of transferring info to LTM rather than the only way = model is detailed, high in validity

Little is known about the central executive - isn't clear how it works and what it does = vagueness means it can be used to explain almost any experimental results, even though it is not scientifically tested

Central Executive can be divided into sub systems - these have not yet been identified and it is difficult in practice to determine which processes control the slave systems that are part of the CE and which are part of other systems = can't explain everything about how memory works. 

Interference Theory - Forget due to info in the LTM becoming confused or disrupted by other info during coding, leading to inaccurate recall

• Proactive - Forwards in time, information stored previously interfers with an attempt to recall something new (memory of an old phone number disrupts attempts to recall a new phone number)
• Retroactive - Backwards in time, coding of new information disrupts information stored previously (the memory of a new phone number disrupts the recall of an old phone number)

Schmidt (2000) - Retroactive interference (no times P's moved to other cities etc), memory of street names learned during childhood

• 700 names selected randomly from a database of 1700 former students of a Dutch elementary school and were sent a questionnaire. 211 P's responeded, age 11- 79 years. 
• Given Molenberg neighbourhood map, 48 street names replaced with numbers, remember as many of them as possible (details collected; times they had moved house, how long lived)
• +ve association with the no of times P's had moved house and street names forgotten, suggesting retroactive interference explains forgetting in some real life situations

Baddeley and Hitch (1977) - Rugby players recall as many teams they had played against, finding the more games played the more forgetful - supporting the idea that interference causes memory loss

Abel and Bauml (2013) - gave P's either single word lists or 2 lists that would cause interference, tested on their memory of them after 12 hours, awake or sleeo, finding that sleep reduced both types of interference = sleep helps strengthen memories, reduce interference, supporting idea that other info interering with memories is the main cause of forgetting

Lab based = not relating to real life, however cause and effect can be established

Forgetting two sets of similar info doesn't happen often = can't explain real life

Doesn't identify the cognitive processes at work = reducing it's clarity, validity and usefulness

• Information is still in the LTM, but can't be accessed
• Recall is dependent upon retrieval cues (labels on files in a filing system)
• Recall is dependent upon accessing info by remembering the retrieval cue under which the info is stored
• Effectiveness depends on how overloaded it is (fewer info, more effective)
• Also how deep the processing of the cue was, how wel the cue fits in with the info
• Two types; Context-dependent failure, and State-dependent failure

Context - External retrieval cures, the external environment is different at recall than coding, fewer marks in a test when sitting in a room not familiar with, than a normal classroom

• Godden and Baddeley (1975) 
• Divers to learn new material on either land or underwater, then tested either on land or underwater
• Recall was worse when it was in a different context to coding (learnt under water, better recalled under water)
• Suggesting being in the same context on learning and recall has a positive effect on recall

Internal retrieval cues, forgetting occuring when an individual's internal environment is dissimilar at recall to when info was coded (trying to recall info learned when sober when you are drunk)

• Overton (1972) - P's to learn info either drunk or sober, tested either drunk or sober
• Recall was worse when P's in a different internal state at recall than at coding 
• Suggesting that being in the same state of learning has a +ve effect on recall

Evaluation of Cue Dependent Forgetting

• Research from Godden and Baddeley, Overton - important role = supporting the idea of cues are essential for memory
• Darley (1973) - Hid money while high, less able to recall again when not high = supports theory as state of marijuana appeared to cause memory loss
• Tulving + Pearlstone (1966) - list of 48 words, (12 categories with 4 words each with headings), those given headings (cue recall condition) recalled more than those without = supporting lack of retrieval leads to forgetting
• Lab based, not every day = can only explain parts of forgetting (lacks ecological validity)
• Smith (1979) - thinking of a cue (classroom), is just as effective as being there = cues can be imaginary (not real like the theory suggests). 

• Jurors often find eyewitness testimony vitally important in making their decision, yet in 75% of cases individuals have been found by DNA wrongly convicted (inaccuarate EWT)
• Bartlett (1932) - memories are not accurate snapshots of events, but reconstructions of events, influenced by schemas (ready made expectations, based on previous experiences, moods, existing knowledge, attitudes and stereotypes).
• Schemas to help us make sense of the world by filling in the gaps, War of the Ghosts, Navojo Indian story, which didn't make sense from their cultural viewpoint, so there memory of the story became distorted, details changed to fit western viewpoings

Factors Affecting EWT

• Misleading Information - Experiences occuring after a witnessed event, a key factor is the use of misleading questions and post event discussion
• Loftus & Palmer (1974) - 1) Speed of Cars, 45 students shown clips, verb either contacted (31.8), hit (34), bumped (38.1), collided (39.3), smashed (40.8) = higher intensity verb, high speed predicted
• 2) 150 P's, clip of car accident, asked with the word smashed, hit or control, if there was smashed glass a week later. Yes/No; Smashed (16/34), Hit (7/43), Control (6/44) = verb effects memory of glass. 

Experiments by Loftus and Palmer = showing clear support that misleading questions/ post event information can lead to distorted memories

Loftus and Pickerill (1995) - 24 P's aged 18-53, 4 stories about their childhood, 3 stories were true, false was being lost in a department store aged 5 and being rescued by and elderly lady. 68% of the true ones were recalled, 29% false ones = false memories can be created from suggestion

Lab setting = lacks ecological validity, doesn't relate to everyday life, however it also means cause and effect can be established

Foster (1994) - EWT was more accurate for real life crimes as opposed to simulations = suggesting low validity, might not apply to real life and therefore the legal court situation as previously thought

One criticism of EWT research is that it often uses artifical scenarios that have no emotional involvement for witnesses. Real life events, such as violent crimes, often have high anxiety content that can greatly effect recall.

Loftus (1979) - Investigate the effect of anxiety on EWT

• P's asked to sit outside a lab where they thought they were hearing general exchanges between people inside the lab
• One condition, heard an amicable discussion about an equipment failure, and a man with greasy hands came out of the lab holding a pen
• In the other condition, they heard a hostile discussion, followed by the sound of breaking glass and overturned furniture, man came out holding a knife covered in blood
• Following the events, P's were given 50 photos and asked to identify the man who had came out of the lab
• Results showed in the scene with the pen, more accurate at recognising the man (49%), than people who had witnessed the scene with the knife (33%)
• Concluding the anxiety caused by seeing the knife, took away attention from the face (weapon focus)

Study from Loftus (1979) - pen/knife = supports anxiety influences EWT as anxiety was present the recall of the crime was reduced

Loftus and Burns (1982) - P's shown a violent version of a crime in which a boy was shot in the face had significantly impared recall for events compared to those who saw a none violent version = showing that the violence, brought anxiety, -ve impact on memory

Deffenbacher (2004) - meta analysis of studies found that high levels of anxiety had a negative impact on the accuracy of crime-related details, including the person = showing that violence (variety) decreases EWT in more ways than one

Lab based - decreasing lab based studies = research from Yuille and Cutshall (1986), interviewed 13 witnesses to an actual violent crime 4 months after the event, and recall was accuarate. However lab based studies have control, so a cause and effect can be established, so we know the anxiety caused the reduced accuaracy 

Christianson + Hubinette (1993) - survey 110 witnesses of bank robberies, the victims who were threatened showed more detailed and accuarate recall than the outlookers = real life incidents involving high stress levels, memory can be accurate, detailed and long lasting

• Suggestions are to improve the ways in which witnesses are questionned by the police, particularly in eliminating leading questions
• Fisher et al (1987) - whether interview techniques need to be improved, over a 4 month period they studied real interviews (standard interviews) conducted by experienced detectives in Florida, finding that witnesses were frequently bombarded with a series of brief, direct and closed questions aimed to elicit facts. 
• Sequence of events were often out of sync with their own mental representation of the event, and they were often interupted and not allowed to talk freely about their experiences (thought these were unhelpful, breaking the concentration of the witnesses, encouraging shorter, less detailed answers. 

Cognitive Interview - Fisher & Geiselman (1992)

• Context Reinstatement (CR) - mentally reinstate the context of the event, recall the scene, weather, thoughts and feelings at the time
• Report Everything (RE) - every detail possible (even if seems trivial/ irrelevant)
• Recall from a Changed Perspective (CP) - describe episode from different points of view
• Recall in Reverse Order (RO) - so event is recalled in different orders

These techniques are all designed to enhance retrieval of the original memory by providing extra cues that may help jog witness' memory for more details

Geiselman and Fisher (1987) - investigate the effectiveness of the cognitive interview, by comparing it to other means used to acquire eyewitness testimony

• 89 P's were shown police training films of simulated violent crimes
• 48 hours later they were interviewed about the films by an experienced Los Angeles police officer, using either the cognitive interview, or the standard, or using hypnosis
• The interviews were recorded and EW reports assessed for correct and incorrect responses
• Results found that the cognitive interview had the most accurate recall, followed by hypnosis then the standard interview. No significant difference
• Suggesting the similar results between the cognitive and hypnosis is due to similar components (report everything). Standard interview due to its repeated focus on encouraging recall of key info without aid to memory retrieval. 

Research led to a version of the cognitive interview called the enhanced cognitive interview. After looking at current police practice through detailed analysis of taped interviewers, Fisher (1998) suggested adding several extra features including 

• The interviewer not distracting the witness with unnecessary interuptions/ questions
• The witness controlling the flow of information
• Asking open ended questions
• Getting the witness to speak slowly
• P's being reminded not to guess and to use the don't know option, when necessary, in order to reduce false memories
• Reducing anxiety in witnesses

Geiselman & Fisher (1997) - greatly improves accuracy = higher validity than SI

Fisher et al (1999) - trained detectives use the enhanced cognitive interview technique with genuine crime witnesses, finding it significantly increased the amount of info recalled = again high validity and usefullness

Modified Cognitive Interview - children, learning difficulties can me interviewed effectively = positive practical applications

Time consuming - more time from officers than they have operational time for = unlikely to be used a great deal

Practical Problem - not generally effective as a form of memory enhancement with regards to recognition of suspects in identity parades and photos = reducing its validiy and usefulness as these methods are still used often with the police