Memory

• What is the short-term memory (STM)?

  > limited-capacity memory store.

  > coding = mainly acoustic

  > capacity = between 5 and 9 items on average

  > duration = about 18 seconds
• What is the long-term memory (LTM)?

  > permanent memory store.

  > coding = mainly semantic (meaning)

  > capacity = unlimited

  > duration = up to life time
• What is coding?

  The format in which information is stored in the various memory stores
• What is capacity?

  The amount of information that can be held in a memory store
• what is duration?

  The length of time information can be held in memory
• Research on coding

  > Baddeley (1966): different list of words to four groups to remember:

  1: acoustically similar (cat, cab, can, etc)

  2: acoustically dissimilar (pit, few, can, etc)

  3: semantically similar (great, large, big, etc)

  4: semantically dissimilar (good, huge, hot, etc)

  > shown original words and asked to recall in correct order.

  > When recalled immediately (STM) - worse with acoustically similar

  > When recalled after time interval (20 mins - LTM) - worse with semantically similar words.

  > suggest: information is coded acoustically in STM, and semantically in LTM.
• Evaluation on research into coding

  (+) identified a clear difference between memory stores: idea of STM uses mostly acoustic coding and LTM mostly semantic has stood test of time, due to support from later research -> important step in our understanding of memory system, which led to the multi-store model.

  (-) artificial stimuli rather than meaningful material: Word lists had no personal meaning to ppts - might not tell us about coding in everyday life. When processing more meaningful information, people may use semantic coding even for STM tasks.

  Suggests: findings have limited application
• Research on capacity

  > Digit span: Jacobs (1887) - read out four digits and ppts recalls these out loud in correct order. If correct, researcher reads out five digits and so on until ppt can't recall order correctly - indicates individuals digit span. found: mean digit span across all ppts = 9.3 items, mean letter span = 7.3

  > Span on memory and chunking: Miller (1956) observed everyday practice - noted things come in sevens. Thought the span of STM = about 7, + or - 2, and people can recall 5 words as easily as can recall 5 letters -> chunking - grouping into units/chunks
• Evaluation of research on capacity

  (+) Jacobs study has been replicated - valid

  Old studies often lacked adequate controls - e.g some ppts digit spans might have been underestimated as were distracted during testing (confounding variable). But, Jacobs findings have been confirmed by other, better controlled studies since

  (-) Millers research might have overestimated STM capacity.

  Cowan (2001) reviewed other research and concluded the capacity of STM is only about 4 (plus/minus 1) chunks.

  Suggests: lower end of Millers estimate (5) is more appropriate than 7 items.
• Research on duration

  > STM: Peterson & Peterson: 4 students, 8 trials - given consonant syllable and 3-digit number. Counted backwards from number until told to stop (vary) - can't mental rehearse. (3,6,9,12,15,18 secs). 3 = about 80% recall, 18 = about 3% - suggests duration about 18 secs

  > LTM: Bahrick: 392 American ppts (17-74 years). Used High school yearbooks. photo-recognition test (50 photos), free recall test (names of graduate class). Photo-recognition: 15 years = about 90% accurate, 48 years = 70%. Free recall: 15 years = 60%, 48 years = 30% -> shows LTM could last up to lifetime.
• Evaluation of research on duration

  (+) Bahrick’s study had high external validity. Investigated meaningful memories (I.e. names and faces). when studies on LTM conducted with meaningless pictures, recall rates lowered.

  Suggests: findings reflect more ‘real’ estimate of duration of LTM

  (-) Peterson and Petersons had artificial stimuli: not completely irrelevant as sometimes have to remember meaningless material (phone numbers) - recalling consonant syllable doesn’t reflect most everyday memory activities where what we try to remember is meaningful.

  Means: study lacks external validity.
• Evaluation of research on duration

  (+) Bahrick’s study had high external validity. Investigated meaningful memories (I.e. names and faces). when studies on LTM conducted with meaningless pictures, recall rates lowered.

  Suggests: findings reflect more ‘real’ estimate of duration of LTM

  (-) Peterson and Petersons had artificial stimuli: not completely irrelevant as sometimes have to remember meaningless material (phone numbers) - recalling consonant syllable doesn’t reflect most everyday memory activities where what we try to remember is meaningful.

  Means: study lacks external validity.
• What is the multi-store model (MSM)?

  A representation of how memory works in terms of three stores called sensory register, short-term memory and long-term memory. Also describes how information is transferred from one store to another, what makes some memories last and what makes some memories disappear.
• What is the sensory register?

  The memory stores for each of our five senses, such as vision (iconic store) and hearing (echoic store). Coding in the iconic sensory register is visual and in the echoic sensory register it is acoustic (ounces). The capacity of sensory registers is huge (millions of receptors) and information lasts for a very short time (less than half a second).
• The multi-store model

  Atkinson and Shiffrin - describes how information flows through memory system. Model suggests the memory is made up of three stores linked by processing.

  
• MSM - sensory register

  > All stimuli from environment pass into sensory register - this comprises several registers, one for each sense. Coding in each store depends of sense - visual = iconic memory, acoustic = echoic.

  > Duration = less than half a second.

  > capacity = high (over one hundred million cells in one eye, each storing data.)

  > Information passes further into memory system only if pay attention to it (attention is key process)
• MSM - STM

  > Coded mainly acoustically and lasts about 18 seconds unless rehearsed - temporary store.

  > Limited-capacity store - only contain certain number of ‘items’ before forgetting - between 5 and 9 items though Cowan’s research suggests it may be 5 rather than 9.

  > Maintenance rehearsal occurs when repeat material to ourselves over and over again. Keep info in STM as long as we rehearse it, if rehearse long enough, passes into LTM.
• MSM - LTM

  > Potentially permanent memory store for info that been rehearsed for prolonged time.

  > Coded mostly semantically.

  > Duration up to a lifetime.

  > Capacity thought to be practically unlimited.

  > According to MSM, when want to recall info from LTM, has to be transferred back to STM through process called retrieval.
• Strength of the multi-store model

  (+) support showing STM and LTM are different: Baddely found we mix up words that sound similar when using STMs. Mix up words with similar meanings when using LTM. These studies show STM and LTM are separate and independent memory stores, as claimed by MSM.

  > counterpoint: Many studies that support MSM used no meaningful material - they used digits and letters (Jacobs), etc

  Means: MSM may not be valid model of how memory works in our everyday lives where we have to remember much more meaningful information.
• Limitations of the multi-store model

  > evidence of more than one STM store: Shallice and Warrington studied KF - STM (digits) very poor when read out loud to him, but much better when read himself. Further studies showed there could be another STM store for non-verbal sounds -> MSM wrong in claiming there’s just one STM store.

  > prolonged rehearsal isn’t needed for transfer to LTM: Craik and Watkins found type of rehearsal is more important than amount. Elaborative rehearsal needed for LT storage - link info to existing knowledge - can be transferred to LTM without prolonged rehearsal.
• What are the three types of long-term memory stores?

  • episodic memory
  • semantic memory
  • procedural memory
• What is the episodic memory?

  Long term memory store for personal events. Includes memories of when the events occurred and of the people, objects, places and behaviours involved. Memories from this store have been retrieved consciously and with effort.
• What is the semantic memory?

  Long term memory store for our knowledge of the world. This includes facts and our knowledge of what words and concepts mean. These memories also need to be recalled deliberately.
• What is the procedural memory?

  Long term memory store for our knowledge of how to do things. This includes our memories of learned skills. We usually recall these memories without making a conscious or deliberate effort.
• Types of long-term memory - research

  > Tulving proposed 3 LTM stores.

  > Episodic: ability to recall events (visit to dentist) - ‘time-stamped’ Store info on how events relate to each other in time -> Make conscious effort to recall. (Aware)

  > Semantic: shared knowledge of world (meaning of words)

  Not ‘time-stamped’. Less personal. less vulnerable to distortion and forgetting than episodic.

  > Procedural: memory for actions or skills (riding bike) - hard to explain to others -> Recall without conscious awareness.
• Strength of types of long-term memory.

  > case studies HM and CW: episodic in both men severely impaired due to brain damage. Semantic relatively unaffected. Procedural also still intact.

  Supports Tulving's view: different memory stores in LTM - one damaged but others unaffected.

  > counterpoint: clinical studies lack control of variables.

  > real-world application: Research has shown memory loss specific to episodic memory. Belleville trained ppts: performed better on test of episodic memory after training than control group - shows treatments can be developed if distinguished
• Limitation of types of LTM

  > conflicting research linking types of LTM to areas of brain: Buckner and Peterson (1996) reviewed evidence regarding locations of semantic and episodic memory. Concluded semantic located in left side of prefrontal cortex and episodic on right.

  However, other research links left prefrontal cortex with episodic and right prefrontal with semantic memories (Tulving 1994)

  Challenges any neurophysiological evidence to support types of memory as there’s poor agreement on where each type might be located.
• What is the Working memory model (WMM)?

  > Baddely and Hitch (1974)

  > A representation of STM. Suggests that STM is a dynamic processor of different types of information using subunits coordinated by a central decision-making system.

  > Concerned with ‘mental space’ that’s active when we’re temporarily storing and manipulating info

  > Consists of 4 main components , each of which qualitatively different especially in terms of coding and capacity.
• WMM - central executive

  > Has ‘supervisory’ role.

  > Monitors incoming data, focuses and divides our limited attention and allocates ‘slave systems’ to tasks.

  > Has very limited processing capacity

  > Doesn’t store information
• WMM - Phonological loop

  > One of the slave systems - deals with auditory information - coding is acoustic.

  > Preserved order in which information arrives.

  > Subdivided into:

  • Phonological store - stores words you hear
  • the articulatory process - allows maintenance rehearsal. The capacity of this ‘loop’ is believed to be two seconds worth of what you can say.
• WMM - Visuo-spatial sketchpad

  > Stores visual and/or spatial information when required (when asked to recall how many windows are in your house you visualise it)

  > Limited capacity - Baddely (2003) suggested its about 3 or 4 objects

  > Subdivided:

  visual cache - stores visual data

  • inner scribe - records the arrangement of objects in visual field.
• WMM - Episodic buffer

  > Added to model by Baddely in 2000.

  > Temporary store for information, intergrading the visual , spatial, and verbal info processed by other stores, and maintaining sense of time sequencing - records events happening.

  > It can be seen as storage component of central executive

  > Limited capacity of about 4 chunks (Baddely 2012)

  > Links working memory to long-term memory and wider cognitive processes such as perception
• Strengths of the working memory model

  > clinical evidence: KF - poor STM ability for auditory info but processes visual info normally. Recall better when he read compared to when read to. phonological loop = damaged, visuo-spatial sketchpad = intact.

  > counterpoint: unclear whether other cognitive impairments (effect performance)

  > dual-task performance support separate existence of slave systems: Baddely - ppts: visual & verbal task at same time: performance similar to when done separately. When both tasks visual or verbal, performance declined - compete for same slave system
• Limitation of working memory model

  > lack of clarity over nature of central executive: Baddely recognised this when he said ‘central executive is most important but least understood component of working memory’. CE needs to be more clearly specified than just being simply ‘attention’ - e.g. some psychologists believe CE may consist of separate subcomponents.

  Means: CE is an unsatisfactory component and this challenged the integrity of the WMM
• What is interference?

  Forgetting because memory blocks another, causing one or both memories to be distorted or forgotten
• What are the two types of interference?

  • proactive
  • retroactive
• What is proactive interference ?

  > Occurs when older memories, already stored, disrupt the recall of newer memories.

  > The degree of forgetting is greater when the memories are similar

  E.g. teacher learned so many names in past, difficulty learning new names
• What is retroactive interference?

  > Occurs when newer memories disrupt the recall of older memories already stored.

  > The degree of forgetting is again greater when the memories are similar.

  E.g. teacher learned so many new names, struggling to remember student names from last year
• Research on interference

  >McGeoch and McDonald (1931)

  > Procedure: studied retroactive interference - changed amount of similarity between two sets of materials. Learned list of 10 words until remember them 100% accuracy. Then learned new list.

  Six groups - different types of new lists:

  1: synonyms

  2: antonyms

  3: words unrelated

  4: consonant syllables

  5: 3 digit numbers

  6: no new list

  > Findings and conclusions: when asked to recall original list of words, most similar material (synonyms) produced worst recall.

  Shows: interference is strongest when memories are similar.
• Explanation of the effects of similarity (interference)

  Reason similarity affects recall - could be due to PI (previously stored info makes new similar info more difficult to store.) or could be due to RI (new info overwrites previous similar memories because of similarity)