exam 1: memory

memory

means by which we retain and draw on past experiences to use info in the present

stages of memory include

• encoding: processes used to store info in memory

• storage: processes used to maintain info in memory

• retrieval: processes used to get info back out of memory

recall tasks

• used to study memory

• asked to generate answers

• includes free recall (any order), serial recall (order matters), and cued recall (cue triggers recall, uses paired associates)

recognition tasks

• used to study memory

• given a choice and asked to see what’s right

explicit memory tasks

involve conscious recollection (recall or recognition); directly retrieving info from memory

implicit memory tasks

require one to complete a task but the performance of the task indirectly indicates memory (ex. brushing your teeth)

procedural memory

remembering and knowing how to do sth

declarative memory

memory for facts or events

models of memory include

• Atkinson & Shiffrin - 3 stage model

• Craik & Lockhart - level of processing

• Baddeley’s working memory model

Atkinson & Shiffrin’s 3 stage model includes

sensory input → sensory memory → attention → short-term memory → encoding (or ← retrieval) long-term memory

sensory memory

• takes info from senses (sensory register)

• fades rapidly and can be pushed out by new info

• keeps only what is processed

Sperling’s test for sensory memory

• whole report procedure vs. partial report procedure

• whole: participants shown a matrix of letters for 50 ms and told to recall max # letters possible; typically remembered ~4 letters

• partial: participants shown matrix for same time but told to report only one row after hearing a tone; able to report any row requested

• proves that attention determines what makes it to the next stage of the 3 stage model

short-term memory

• maintained by rehearsal

• limited capacity, thought to be 7 plus/minus 2

• chunking increases effective capacity

Peterson & Peterson (1959) tests for short-term memory

• subjects shown list of non-sense words

• given a distractor task

• cue was given and subjects told to recall the words

chunking

• grouping items together reduces the number of items by redefining them in another meaning

• ex. to remember a list of numbers: grouping them in 4 like years, numbers grouped as a birthday, numbers combine to be a race time, etc.

does expertise help with chunking?

yes; compared chess masters to novices by showing board for 5 sec to memorize arrangement and asked to recreate on an empty board → chess masters were better when original board contained actual game positions

long-term memory

• stable

• can be retrieved and activated again in STM

• limitless duration, limitless capacity

Bahrick’s tests on long-term memory

• 392 ex-high schoolers 17 - 74 took 4 tests using HS yearbooks: photo recognition, name recognition, name & photo matching, free recall of names of classmates

• found that there was 90% accuracy for face and name recognition after 34 years; LTM is quite stable

Craik & Lockhart’s level of processing

• different ways of processing info will lead to different strength of memories

• deep processing: better memory, elaborates according to meaning, elaborative rehearsal

• shallow processing: emphasizes physical features, quickly decays, maintenance rehearsal

Craik & Watkins (1973) test on level of processing

• participants listened to list of words

• asked to recall last word in list beginning w/ a certain letter (ex. “t”)

• varied # of intervening words (ex. words that don’t begin w/ “t”) to vary time in STM

• found that recall of words was independent of length of time in STM/# intervening words, so maintenance rehearsal did not automatically lead to LTM; students rehearsed w/out elaborating on meaning (shallow processing)

Craik & Tulving (1975)

• participants studied word list in three different ways: structural (capital letters?), phonemic (rhymes with dog?), semantic (see if word fits in sentence)

• given recognition test (was word in the study list?) to see which type of processing led to best memory

• found that there was higher recognition for semantic and rhyme than structural (more meaning → deeper)

self-reference effect

• if subjects see words in a list that describe them, they have a higher level of recall

• level of recall was even higher if the word was positive and that they identified with

criticisms of level of processing model (Morris, Bransford, and Franks 1977)

• memory performance also depends on match between encoding processes and type of test

• participants given two processing tasks (semantic and rhyme) + two types of tests (recognition and rhyme)

• based on LOP only, you’d expect rhyme to do better on both b/c it involves deeper processing

• however, rhyme processing only did better than semantic when rhyme test was used; semantic processing did better than rhyme when recognition test was used

Baddeley’s working memory model

• focuses on importance of working memory in modern cognitive psychology

• involves central executive, visuo-spatial sketch pad, phonological store

phonological/articulatory loop

part of working memory used to maintain verbal (and some auditory) info for a short time

word length effect

• applies to Baddeley’s working memory model and phonological lop

• subjects remember short words better than longer words

articulatory suppression

• applies to Baddeley’s working memory model and phonological loop

• repeatedly saying a distractor word while viewing a list of words interferes with the loop and decreases rehearsal

visuo-spatial buffer

part of working memory dedicated to visual imagery and spatial processing

info enters the visuo-spatial buffer from

visual perception or from long-term memory

dual-task (Brooks interference tasks)

• applies to Baddeley’s working memory model

• when doing 2 tasks involving the same loop, performance will degrade: ex. tracing the block letter F while having to point to yes/no when asked if words in a sentence are nouns

• able to complete 2 tasks involving different loops with more efficiency: ex. tracing the block letter F while having to say “yes”/“no” when asked if words in a sentence are nouns

central executive

part of working memory that

• supervises attention

• plans and coordinates activities

• monitors mental activities, checks on content of buffers

• pulls info out of LTM to use in working memory

conductor Toscanini and “S”

• success of memory, known for extraordinary memories

• Toscanini remembered ~250 symphonic works and words/music for ~100 operas

• S was a newspaper reporter that never took notes and remembered long lists and passages; could repeat 70-80 word lists

sensory synesthesia

• unusual and usually involuntary associations between different sensory modalities or representations

• often a tool for mnemonists

are normal people capable of sensory synesthesia? (Shepard 1967)

• yes!

• showed people pics of 612 common objects → tested to see if participants could recognize if they saw the objects or not

• accuracy was generally high but declined after time passed (timescale of many days)

Hermann Ebbinghaus & non-sense syllables

• memorized nonsense syllables until he had 100% accuracy

• Ebbinghaus forgetting curve - words had no meaning so there was a rapid decrease in retention

why do we forget?

• often failed to completely encode in the first place

• shallow depth processing creates weaker memories

factors improving retention/recall

• spacing of learning

• organization of info

• state/context dependent memory

spacing of learning improves retention and recall because

• paying less attention with a longer time is less effective than paying more attention for spaced and shorter intervals

• studying on different occasions could mean learning in more contexts → more routes to retrieval

• richer encoding scheme due to focusing on different aspects of material each time

proof for organization of info improving memory

• Bower et al (1969)

• hierarchical organization of place names into countries, cities, etc. improved memory over the unorganized list

• 19% accuracy → 60% accuracy when organized

state and context dependent memory improve learning and retention by

matching the context of learning and memory

• Godden & Baddeley: learning words underwater or on the surface in a classroom → tested underwater or in classroom - same contexts improved learning

• Grant et al: studying in quiet vs w/ cafeteria noise recording → testing - matching contexts improved accuracy

• time of day also matters; learning at 3 pm → perform better at 3 pm

context-dependent

external, environmental factors

state-dependent

internal, physiological factors

questions on LTM: if you can’t retrieve a memory from LTM, has it disappeared?

“It’s still there” theory:

• Nelson (1971)

• gave subjects a paired associate list and tested with cued recall (43-house → asked to fill in blank: 43-__)

• after 2 weeks, subjects recalled 75% of the list; then had to focus on the rest that they’d forgotten

• some subjects relearned the original pair or were taught a different association (ex. 43-house or 43-kettle)

• those that were taught the different association had a lower rate of relearning vs. relearning the same association from before (43-house → 78% relearned, 43-kettle → 43%)

• indicates that some memory was left for “forgotten items”

questions on LTM: if you can’t retrieve a memory from LTM, has it disappeared?

“It’s forgotten” theories:

• decay theory: memory weakened with disuse and passage of time

• interference theory: constantly learning new things that interferes with what was previously known

• control and experimental group

  • control: learns list A → delay → tested for A

  • experimental: learns list A → learns list B → delay → tested for A → lower retention

interference theory: interference can either be

• proactive: old memories interfere with newer info

• retroactive: new memories interfere with recall of older info

anterograde amnesia

affects memory after trauma

retrograde amnesia

affects memory before trauma

role of hippocampus in memory

• memories are not permanently stored there

• but responsible for encoding, consolidating, rehearsing info

• eventually stored in the cortex and “binded”

mirror-reading study: normal subjects vs. amnesia patients

• tried to learn mirrored words for 4 days; some words are repeated out loud (implicit + explicit memory) while others aren’t (only implicit)

• both groups improve learning for non-repeated words

• only control group does better with repeated words

• amnesia patients struggled with retaining explicit memory but not implicit

eyewitness testimony can easily become an example of

• faulty memory

• memories are easily distorted in a number of ways: forgetting, remembering things that didn’t happen, being influenced easily, susceptible to misinformation

misinformation effect & Loftus experiments

misinformation incorporated after the event impacts memory

• subjects would see an event

• undergo questioning where misinformation was presented (often as a question or different framing/wording)

• memory tested to see if it was affected by misinformation → memory was often distorted

• ex. “smashed” about a car crash → 41 mph estimated, “contacted” → 32 mph

can false memories be created?

• yes!

• college students given list of their life events + fake memorable events created by researchers

• had good memory for real events, could not recall fake events at first but became real to participants after a few asks

• some participants were talked into new memories

factors affecting acceptance of false memories

• plausible events easier to implant

• repetition of false info

• when subjects are told to imagine it, it increases belief

• some people are more susceptible to believing