Cognitive Psychology: Memory

• Duration means how long a memory can last before it is no longer available
• Duration of LTM is unlimited, duration of STM is measured in seconds

Research

• Shepard (1967) - showed participants photos. After 1 hour they showed almost perfect recognition. After 4 months they recalled about 50% of the photos.
• Bahrick et al. (1975) - recall yearbook photos. After 48 years they were around 70% accurate
• Peterson and Peterson (1959) - gave participants consonant syllables (non-meaningful) followed by a 3-digit number. Participants had to count back from number in 3s or 4s for  3, 6, 9, 12, 15 or 18 seconds  before recalling the syllable. After 3 seconds: 90%. After 18 seconds: 2%.

Evaluation

• Marsh et al. (1997) - duration may only be 2 seconds if participants don't expect to recall info
• Nairne et al. (1999) - duration could be 96 seconds if the same items were being recalled across the trials

Validity

• Only one kind of memory (semantic) was being tested
• Not actually testing duration, instead they were testing displacement

• Capacity means the amount of information that can be stored
• In LTM capacity is unlimited, in STM it is less than 7 chunks

Research

• Miller (1956) - digit span of 7 +/- 2 in STM
• Simon (1974) - size of chunk matters - 8-word phrases meant a shorter memory span than one-syllable words

Evaluation

• Cowan - 4 rather than 7 chunks

Individual Differences

• Jacobs (1887) - suggested age differences in digit-span

Applications

• Postcodes are based on chunking

• Encoding means the way that information is changed so that it can be stored in memory
• In LTM encoding is semantic (meaning), in STM it's acoustic or visual

Research

• Baddeley (1966a and 1966b) - acoustic similarity decreases STM recall, semantic similarity decreases LTM recall

Evaluation

• Brandimote et al. (1992) - visual encoding in STM if given visual task and verbal rehearsal prevented
• Frost (1972) - visual coding in LTM, not just semantic

Research

Structure:

• Sensory memory - evidence from Sperling (1960) - when participants were asked to recall all 12 items they remembered 5 items (42% accuracy). When asked to recall one row of 4 items, they remembered 3 (75%)
• STM - limited capacity and duration, mainly acoustic coding
• LTM - unlimited capacity and duration, mainly semantic encoding
• Distinction between STM and LTM - Glanzer and Cunitz (1966) - gave participants a list of 20 words which they then had to recall. They remembered words from the start of the list (primacy effect) and the end of the list (recency effect) but forgot words from the middle.
• Primacy effect occurs because words at the start are best rehearsed and transferred to LTM. Recency effect occurs because words at the end are still in STM when people are recalling.

Processes:

• Attention and maintenance rehearsal - need to pay attention so the information can be encoding. Need to rehearse information in order for it to be transferred from STM to LTM. 

Strengths

• Strong evidence to support claims for duration, capacity and encoding
• Includes details of structure process
• Has stimulated a lot of research which leads to increased understanding
• Serial position effect (Glanzer and Cunitz), role of hippocampus (Squire et al., 1992), case studies of brain damage (e.g. HM)
• Evidence for the three stores: see above

Limitations

• Oversimplified
• STM doesn't function as a unitary store, e.g. KF (Shallice and Warrington, 1970) - had brain damage which meant difficulty with verbal information in STM but no problem with processing visual information
• LTM is not a unitary store: e.g. semantic, episodic, procedural memories
• Processing more important than maintenance rehearsal - Craik and Lockhart (1972) compared shallow, phonemic and semantic processing - concluded that things which are processed more deeply are more memorable because of the way they're processed
• STM is not independent of LTM - Ruchkin et al. (2003) compared words and pseudo-words involved different brain activity

Applications

• Guiding rehabilitation of people with hippocampal damage

Validity

• Memory research usually involves semantic memory - relates to some everyday memory activities (remembering a phone number) but not all aspects.
• Participants often psychology students - might try and find out what the experiment is about, affecting their behaviour
• Lab experiments (demand characteristics, experimenter bias but well controlled)

Research

• Explains why you can do two different tasks at the same time but not two similar tasks (evidence from Baddeley and Hitch)
• Central executive - resource allocation, small capacity
• Phonological loop - phonological store and articulatory process, maintenance rehearsal. Deals with auditory and speech-based information
• Visuo-spatial sketchpad - visual cache and inner scribe. Deals with locations and movement
• Episodic buffer - general store and integrates information from other stores as well as from LTM

Applications

• WMM useful in diagnosis of mental illness

Validity

• Same validity issues as for MSM.

Strengths

• fMRI shows central executive more active when two tasks done (Bunge et al., 2000)
• Explains word-length effect (Baddeley et al., 1975a) - people can cope better with short words than long words in working memory (STM)
• Visuo-spatial sketchpad shown in visual tracking task (Baddeley et al., 1975b) - participants had to track a moving light with a pointer while either describing the angles on the letter F or performing a verbal task. Verbal task and tracking was easier, because they involved two different systems
• Episodic buffer shown comparing recall of related and unrelated words (Baddeley et al., 1987) - recall was much better for sentences (related words) than for unrelated words.
• Explains memory deficits of KF and SC - both their phonological loops were affected by brain damage but they had no trouble with anything else.
• Emphasises process rather than structure, unlike MSM.

Limitations

• Central executive doesn't actually explain anything
• Central executive is probably several components (Eslinger and Damasio, 1985) - EVR did well in reasoning tests, but struggled with decision-making.
• Problems with evidence from brain-damaged individuals, e.g. no before and after comparisons

Research

• Loftus and Palmer (1978) - experiment 1, misleading information (hit vs smashed), led to inaccurate recall
• Experiment 2 (broken glass), misleading information changes information storage

Evaluation

• Supporting evidence from Loftus et al. (1978), stop/yield sign
• Challenged by Bekerian and Bowers (1973) - misleading information affects retrieval
• Real-life robbery (Yuille and Cutshall, 1986) - accurate recall despite misleading information

Individual Differences

• Males and females same (Wells and Olsen, 2003)
• Older people have more difficulty remembering source of information but recall for event same as younger people (Schachter et al., 1991)

Applications

• Mistaken EWT largest single factor in wrongful conviction

Validity

• Lab experiments may not be taken seriously
• Participants not emotionally involved
• Foster et al. (1994) - better identification with real-life set up

Research

• Children make more errors (Parker and Carranza, 1989)
• Younger participants more confident but no age differences (Yarmey, 1993)
• Older less accurate when delay was one week (Memon et al, 2003)

Evaluation

• Own-age bias (Anastasi and Rhodes, 2006)
• Due to differential experiences

Individual Differences

• Alcohol impairs attention (Clifasefi et al., 2006)

Research

• Deffenbacher et al. (2004) meta-analysis showed anxiety reduced accuracy of EWT
• Christianson and Hubinetter (1993) anxiety increased accuracy in real-life bank robberies
• Weapon-focus effect - Johnson and Scott (1976) - if a weapon is used participants are more likely to focus on that rather than the person holding it

Evaluation

• Contradiction explained by Yerkes-Dodson law
• Meta-analysis (Stebly, 1992) supports weapon-focus effect
• Loftus et al. (1987) tracked eye movements to support weapon-focus effect

Applications

• Riniolo et al. (2003) - EWT from Titanic was accurate

What is it?

• Report everything
• Re-create original context
• Change order
• Change perspective

(Fisher and Geiselman, 1992)

Evaluation

• Milne and Bull (2002) - report everything and mental reinstatement gave best recall
• Hard to evaluate because there are many versions of the CI
• Time-consuming
• Enhanced CI creates greater demand on interviews
• Kohnken et al. (1999) - meta-analysis, 34% increase in correct recall

Applications

• CI used in Brazil (Stein and Memon, 2006)

Individual Differences

• CI works better than SI (Structured Interview)

Mnemonic Techniques

• Verbal - acronym, acrostic, rhymes, chunking
• Visual imagery - method of loci, keyword method (Atkinson and Raugh, 1975), spider diagrams and mind maps

Research Evidence

• Verbal mnemonics - popular with students; effective in children with learning disabilities (Glidden et al., 1983)
• Visual imagery mnemonics - method of loci useful for older adults (O'Hara et al., 2007)

Explaining How They Work

• Organisation - creating memory hooks and organising material (Bower et al., 1969)
• Elaborative rehearsal - amount and nature of rehearsal
• Dual coding hypothesis - Paivio (1971)

Applications

• Mnemonics used to overcome STM deficits in Down Syndrome children (Broadly and MacDonald, 1993)

Limitations

• Most research in laboratory conditions rather than real-life conditions (Slavin, 2005)