Human Memory Midterm 1
2 Main Uses for STM
1) Brief holding place for new information while we are learning it e.g.: memorizing someone's phone number 2) A place where information is processed e.g.: doing a math problem in your head (23x7 = ?) e.g.: understanding a sentence
Two Types of Visual STM
1) Spatial STM = where things are 2) Object STM = where things are
Phonological Loop
= A model of verbal short term memory store Composed of two parts: 1) A memory store that holds information: Short Term Phonological Memory store that is limited in capacity The items are registered in the store as memory traces that decay within a few seconds, but can be refreshed by rehearsal of material. 2) Articulatory Rehearsal Process: A system by which we rehearse material vocally or sub-vocally
Articulatory Rehearsal Process
= A system by which we rehearse material vocally or sub-vocally to help us access + strengthen memory traces The longer it takes for us to rehearse/articulate, the worse our memory gets
Word Length Effect
= A tendency for verbal memory span to decrease when longer words are used (e.g.: words with more syllables) Baddeley et al (1975): Subjects recalled lists of 5 words in correct order -> As the number of words increased, performance declined List of one syllable words -> 90% accuracy List of five syllable words -> 50% accuracy
Theories of Memory Structure
Historically, theories of memory structure have related memory to whatever current technologies they had
Short Term Forgetting
Preventing a subject from thinking about or otherwise rehearsing a sequence in STM causes the subject to forget the information
Phonological Loop & the Phonological Similarity Effect
Difficulty in recalling similar sounding items occur when trying to access memory traces Similar items have fewer distinguishing factors than dissimilar items The articulatory rehearsal process can be blocked if the subject is required to continually say some piece of irrelevant material (e.g.: repeating the word "the") Thus items are kept from entering the phonological store because the subject cannot sub-vocally name the visually presented item
Sperling + Partial Report Technique
Flashed the matrix for 1/20th of a second Immediately afterward, different tones were sounded to indicate which row to report Subjects did not know which row they were going to be required to report until the matrix disappeared RESULTS: 1) Subjects remembered almost 100% of the 9-letter matrix and about 75% of the 12-letter matrix 2) Suggests subjects have approximately 9 letters available in sensory memory, which is very different from the 3 or 4 letters recalled with the whole report procedure 3) Suggests that all information in the matrix is available immediately after the disappearance of the stimulus 4) The difference in performance between whole- and partial-report suggests that the information in visual sensory memory decays very quickly
Explicit Memory
Memory of facts and experiences that a person can consciously know and state / declare that we know e.g.: state capital, episodic memory, semantic memory Also referred to as declarative memory Explicit memory can be stated, while implicit memory has to be demonstrated
Scanning
Occurs at the rate of about 1 letter every 10 ms, according to Sperling
Chunking
Organizing information into larger, more manageable and familiar units in order to improve memory span A "chunk" can be a word, phrase, sentence or any other information that we can put together to assign a single meaning or idea
Keppel and Underwood (1962)
Performed typical Peterson-type experiments Subjects performed a number of similar recall trials Each succeeding trial is subject to interference from the preceding trials Only the first trial was free from the interference Data showed NO effect of rehearsal-prevention interval on memory for the first sequences subjects were shown Subjects had trouble discriminating b/w the current sequence and the earlier sequences
Short Term Forgetting (Peterson and Peterson)
Presented subjects with sequences of 3 unrelated consonants Sequences were shown for 3 seconds Immediately afterwards, subjects are shown a 3 digit number and were instructed to count backward by 3's up to 18 seconds (rehearsal prevention task) Subjects were the instructed to recall original 3 consonants Subjects then performed more similar trials
Serial Recall Task
Subject is given a long list to recall but they have to recall the items in the order presented
Partial Report Technique
Subject only has to report PART of the material
Spatial Short Term Memory (Posner and Konick 1966)
Subjects were asked to remember the point along a line at which a circle was presented for 1 second For varying periods of time between presentation of the stimulus and attempt to recall, subjects would perform an intervening (non-spatial, non-visual) task Subjects were asked to recall the locations of circles after 0 seconds, 5 seconds, 10 seconds or 20 seconds The intervening tasks were of various difficulties: 1) Rest (no intervening task) 2) Recording digits pairs onto a sheet of paper 3) Adding digit pairs and writing them onto a sheet of paper 4) Classifying digit pairs as being high or low, and even or odd RESULTS: All the intervening tasks affected performance The more difficult and demanding the task, the more performance was affected This was particularly true for longer delay periods Implies that if we want to maintain things in spatial STM, rehearsal is required even if the task is non-visual and non-spatial Subjects said it became difficult to recover the mental "picture" as the intervening tasks increased in difficulty
Why are the Corsi and visual span tests thought to be good measures of spatial and object memory? (Della Salla 1999)
Subjects were given the Corsi/ Visual patterns test with an intervening task For both the Corsi and Visual patterns test, subjecs performed different types of intervening tasks between stimulus presentation and time of recall: 1) Spatial intervening task 2) Visual intervening task RESULTS: A person's performance on the Corsi Task was impaired much more by the spatial task, than the visual intervening task The visual pattern test performance was disrupted much more by the visual intervening task than by the spatial intervening task
Articulatory Suppression
The articulatory rehearsal process can be blocked if the subject is required to continually say some piece of irrelevant material (e.g.: repeating the word "the") Prevents the subject from sub-vocally naming visually presented items, keeping them from entering the phonological store
Sperling's Experiments (1960)
The experiments investigated the nature of sensory memory An array of letters was flashed for 1/20th of a second [K Z R] [Q B T] [S G N] Results: Subjects could only report 4 or 5 letters Each subject's immediate memory span was basically constant, no matter how many letters were in the stimulus It also didn't matter how long the stimulus was shown for (between 15 ms and 500 ms)
Memory Span
The longest list of items a person can accurately report back (in correct order) 50% of the time The list of items can include digits, letters or words
Atkinson Shriffin Model
The model is a metaphor: the 3 memory stores (sensory, STM and LTM) are not viewed as actual anatomical structures in the brain, but as functionally distinct "types" of memory Three types of memory: sensory memory, short term memory, and long term memory Sensory memory can transfer into STM through attention, and STM can transfer into LTM through rehearsal
Digit Span
The number of digits that a person can retain and recall It is a specific type of memory span 1) Subject is presented with a sequence of digits 2) Subject attempts to repeat back the digits in the order they were presented 3) Lengths of sequence increased until the subject fails
Example of Phonological Loop: Digit Span
The sequence of digits you have to recall are placed in a phonological store In order to keep the memory trace accessible, we articulate the rehearsal process HOWEVER, if the sequence of digits is too long, it takes too long to rehearse he entire sequence and increases chances that the memory traces may fade away before being refreshed The rehearsal process is dependent on how long it takes someone to articulate the information Digit span is also affected by the language spoken by the subject: Chinese language can articulate digits in their language quickly, which is why they have a longer digit span. The longer it takes for us to rehearse/articulate, the worse our memory traces.
The Visual-Spatial Distinction
There are some standardized tests that psychologists have established to measure spatial STM vs object memory performance: 1) Corsi Block Tapping Test 2) Visual Patterns Test
Spatial Intervening Task
Touch and follow a series of pegs stuck in a peg board with the pegboard covered and out of view No visual elements
Object Memory (Vogel, Woodman & Luck 2001)
Tried to find out how many objects we can maintain in STM by showing subjects bars with different features (e.g.: color, orientation size, texture) Subjects would be shown an array of these stimuli for 100ms, and then they would see another array of these bars 900 ms later The task was for the subject to decide if the 1st and 2nd presentations of bars were the same (half of the experiments had different bars, the other half had the same) RESULTS: subjects were 90% accurate in identifying when a single feature had changed in a single stimulus BUT after 4 objects, there was a drop off Our ability to recall does not seem to be influenced by how complex the objects are - what is more important is how many objects we are trying to maintain Subjects were just as good at identifying a change in one of four features as identifying a change in a given feature
Irrelevant Speech Effect
= Exposure to speech sounds (even irrelevant speech sounds like nonsense syllables or foreign words) impairs subjects' memory for visually presented lists of digits Occurs even when subjects are told to ignore the irrelevant speech Braddeley's interpretation: Speech automatically enters the phonological store, even without us paying attention This irrelevant speech ends up adding noise to the phonological memory trace of the list items being rehearsed, making it more difficult to accurately recall list items
Decay
= Forgetting due to disuse of information E.g.: failing to rehearse the information
Improving Memory Span: Rhythmic Grouping of Items
= Presenting items with a particular rhythm Grouping items in three's (leaving a small gap in between groups) is optimal, and grouping items in two's is the next best e.g.: phone numbers 949-824-15-00 Also useful in long term remembering as well e.g.: AC Aitken remembered first 1,000 digits of Pi by arranging digits into groups of 5 digits, reading them over and over in a particular rhythm
Phonological Similarity Effect
= The tendency for immediate serial recall of verbal material to be reduced when items are similar in sound = Impaired memory for items in a list that sound similar E.g.: it is harder to remember (cat, map, man, cap, mad) correctly than to remember (pen, day, sup, cow, hot) correctly Effect is due to similarity in sound, not similarity in meaning (there is a small effect for words w/ similar meanings, but it is not nearly as drastic as words with similar sounds) May suggest that we store things in STM based on sound
Interference
= When learning of some material decreases recall of other material Repeating semantic categories can cause interference
Visual Span
= the largest number of matrix cells that can accurately be recalled is about 16 cells
Corsi Span
= the largest series of blocks a subject can typically recall from a Corsi Block Tapping Test is about 5 blocks
Short Term Phonological Memory Store
A memory store that is limited in capacity Items are registered as memory traces that decay within seconds, but can be refreshed by rehearsal of material
Wax Tablet Analogy
A theory of memory structure that the Ancient Greeks used Memory is like a block of wax that differs in "size" and "hardness" for different individuals Some people have a big wax tablet (can hold a lot of information), others have small Hard wax tablet means that it is hard to store information into memory Remembering was like stamping an impression into the wax - as long as the image remained in the wax, so did the memory remain intact
Visual Image Store
AKA iconic memory
Rehearsal Prevention Task
An activity that prohibits the repeating of test items
Sperling's Conceptualization of Visual Sensory Memory
An image immediately goes into a memory store that Sperling named the Visual Information Storage (VIS), but it starts to delay fairly rapidly The information is available to be reported through a process known as 'scanning' Scanning occurs at the rate of about 1 letter every 10ms To demonstrate this, Sperling used a masking procedure: 1) Pre-field: either dark or "noise" made up of little scattered bits of letters 2) Post-field: always noise RESULTS: With dark pre-field: subjects could report an additional letter about every 10 ms With "noise" pre-field: there was an initial delay in scanning, but then they scanned a letter every 10 ms
Sensory Memory
Automatically registers any environmental stimulus to hit the sensory receptors Only holds information for a few seconds, or fractions of a second after the actual sensory stimulation is over
Clive Wearing
Classical musician Viral infection caused encephalitis, resulting in major brain damage Had severe impairment of memory with no ability to form new long term memories Constantly believes he has just returned to consciousness, but has retained some memories from his earlier life Although his general knowledge is greatly affected, his musical skills (motor skills) remain in tact Demonstrated that our memory is not just a single system (because if it was, then all his memories would have been gone)
Duration of Visual Sensory Memory (Sperling, 1963)
Delayed presentation of indicator tone Hypothesis: the longer the delay before the indicator tone, the more visual sensory memory decays, and the poorer the subject will perform Results: Performance declined as the length of the delay of the indicator tone increased
Release from Proactive Interference
Hypothetical situation: the words come from the same semantic category, not random (e.g.: animals, fruit, metals) Typical results: people do well on the first sequence then do worse until there is a change in category The change in categories lets you avoid confusing earlier words with current sequence Release from proactive interference: changing the category of presented words resulted in improved recall because when the semantic category changes, people's performances on the first trial immediately improves again Works with long term memory as well REVERSE: Repeating categories CAUSES interference
Example of Sensory Memory: A sparkler trail
If you move a sparkler in a circle quickly enough, you'll see a complete circle Sensory memory preserves the image long enough to perceive a continuous circle rather than separate points of light
Primacy Effect
Improved memory for items at the beginning of a list Can be explained because words at the beginning of the list must not be in STM, since they are remembered even when there is a rehearsal prevention task Perhaps having more time to rehearse these items at the beginning of the list encoded them into LTM Glanzer and Cunitz (1966): Presented list at a slower pace, allowing subjects more time to rehearse, leading to an increased memory for words at the beginning of the list
Encephalitis
Inflammation of the brain that affected Clive Wearing and resulted in severe brain damage
Proactive Interference
Interference in recall caused by previous exposure to some material
Semantic Memory
Knowledge about the world (facts, rule, concepts) stored in long term memory
Stephen Wiltshire
Known as the Living Camera Diagnosed with autism at age 3 Developed an interest in drawing @ age 4 Savant: could look at things even briefly and later draw them in great detail Can draw entire cities after a brief helicopter ride around the city
Long Term Memory
Limitless storehouse of information Stores information for longer periods of time Limitless capacity or duration Examples: conversations from half an hour ago, events from our childhood, capital of Missouri Different types of LTM: episodic, semantic, implicit, explicit
Episodic Memory
Long term memory for particular events / episodes in your life
Visual Intervening Task
Looking at pictures of abstract paintings No spatial elements
Priming + Amnesia
Priming effects are distinct from conscious awareness because they are an implicit memory Amnesia patients will show priming effects (e.g.: if you show an amnesiac a Where's Waldo picture, they can quickly find Waldo again later, even if they can't recall ever seeing the picture before) Amnesiac patients can also learn things without being aware that they've learnt them because implicit memory retention is independent of any conscious recollection
Murdock (1961)
Rather than use 3 unrelated letters like Peterson and Peterson, they used 3 letter words (e.g.: C-A-T) One condition: 3 different 3 letter words Second condition: 1 3-letter word People's performance with 3 words is almost the same as using 3 random alphabet letters Shows that it really is more about the things rather than individual letters because of chunking People did very well in he 1 word, 3 letters condition
Information Processing Theory
Relates the mind to the computer in the way that information is stored in memory Both computers and mind analyze and transform information Many different types of models, but most divided memory into 3 phases: 1) encoding, 2) storage, 3) retrieval Most influential model was the Atkinson-Shriffin model
Rehearsal
Repeatedly verbalizing or thinking about information Enough rehearsal can move something from short term memory into long term memory
Decay vs. Interference
Results of Peterson & Peterson and Murdock (1961) originally thought to be due to decay Interference is also known to affect recall To accept the decay explanation, you have to assume that there is no interference present in the task 1 possibility is that the rehearsal prevention task is a source of interference Later, studies showed that there was actually a different source of interference in Peterson-type experiments
Short Term Memory
Retains information only temporarily Often referred to as working memory
Implicit Memory
Retention independent of any conscious recollection or awareness Things that we can't really declare, but we can demonstrate E.g.: motor skills (typing, musical ability) Happens outside conscious awareness Implicit & explicit memory seem to be handled by different systems: Amnesiac patients can't recall facts or what they've done recently, but they can retain certain things related to implicit memory (e.g.: Clive Wearing) Learning that is demonstrated by priming effects is also an implicit memory Measurement of implicit memory is done indirectly (seeing how long it takes an amnesiac to find Waldo instead of just asking them where he is in a picture they have already seen)
Echoic Memory
Sensory memory for hearing Lasts up to 3 or 4 seconds Limited to one or two items - contents of echoic memory can be "wiped out" or "over written" by additional items Even if you weren't paying attention to an auditory stimulus at first, you can "replay" it and still have access to it as long as there is no new stimuli Shows recency effect BUT if another spoken item is inserted between the last item on the list and time of recall, that "recency" advantage goes away Geared towards speech: anything without verbal content does not seem to affect the recency effect Longer lasting than iconic memory Can be overwritten by presenting further irrelevant items Spoken items between the end of the list and point of recall disrupt echoic memory, but other sounds (e.g.: buzzer, tones) do not
Sensory Receptors
Specialized cells that respond to physical energy from the environment Once the receptor cells have been activated, the record of this activation is stored in sensory memory
Corsi Block Tapping Test
Standard measure of spatial memory Array of 9 blocks laid out Experimenter taps a particular sequence of blocks, and subject is asked to repeat (tap) the same sequence Subjects begin with 2-block trails and proceed to longer sequences until subjects can no longer remember Corsi Span = the largest series of blocks a subject can typically remember is about 5 blocks
Capacity of STM (George Miller)
Studied memory span Discovered that on average, most people can remember about 7+/- 2 things in their short term memory There are exceptions: some people can remember as many as 10 things, some people as little as 4
Whole Report Technique
Subject is asked to recall ALL the material Used for Sperling's first version of the experiment when he flashed the array and asked subjects to report back whatever they could
Free Recall Task
Subject is given a long list of items to recall as many items as possible in any order they want
Are people using verbal STM to help remember object features? (Vogel et al 2001)
Trying to control for the question if subjects are verbally repeating information to help them remember features instead of using visual STM Vogel et al employed articulatory suppression to prevent verbal rehearsal in order to investigate If people were really using verbal STM, there should be a drop in performance due to the articulatory suppression RESULTS: there was no change in performance; people do not seem to be using verbal STM to "verbalize" features of he stimuli
Visual Patterns Test
Used to measure object memory Subject is shown a matrix of cells; the matrices are of different sizes with different numbers of cells Half of the cells are filled in, and the other half of the cells are blank The subject is shown a particular pattern and asked to reproduce it on a blank matrix Visual span ends up being about 16 cells
Masking Effects in Visual Sensory Memory (Sperling, 1963)
Varied brightness of pre-field and post-field (could could be in darkness or in brightness before or after presentation of stimulus) Investigated masking effect: how does the stimulus being presented just before or after the matrix of letters affect the duration of the memory store? RESULTS: 1) Regardless of brightness of pre-field and post-field, performance declined as the length of the delay of the indicator increased 2) Brightness of pre-field and post-field affected how rapidly performance declined 3) When pre-field and post-field are bright, performance declined very rapidly When there was a half second delay, performance was no better than with the whole report method 4) However, when pre field and post field were dark, performance declined more slowly It took about 4 or 5 seconds of delay of the indicator before it reached the performance level found with the whole report method
How Items Get Stored in the Phonological Store
Verbal material we hear is automatically placed in the phonological store, but visually presented items make it into the store only if they are things we can name via vocal or sub-vocal articulation E.g.: seeing a car, saying the word car
2 Systems of Short Term Memory
Visual STM and Verbal STM
Iconic Memory
Visual information storage later became known as iconic memory Hold visual information until it can be reported or transferred to a more durable, longer-lasting non-visual memory system through scanning Duration of this visual memory store is affected by a bright field - it causes information in iconic memory to decay much more rapidly Disruptive mask delayed when the scanning would start but did not affect how long it took
Ability to recall objects over varying time intervals
Vogel and team altered the time interval to test subjects ability to recall objects over varying time intervals They looked at subjects' ability to recall after 900ms delay, 2900ms delay, and 4900 delay RESULTS: there was only a very slight, non significant drop
Sperling (part 2)
Wanted to investigate if subjects stored the entire array briefly, even though they could only report 4 or 5 of the letters Argued that we do store the entire image/array briefly, but it fades very quickly, preventing subjects from demonstrating this memory To test his theory, he needed to come up with a way to demonstrate that subjects do see the entire array, but without asking them to report the entire array Thus, he used the partial report technique
Priming
When an object has been perceived or process, it's the tendency for that object to be perceived more easily next time we're exposed to that object It is a temporary effect E.g.: after recently reading the word "rabbit", it becomes easier to perceive the word "rabbit" if that word is briefly flashed, and you are more likely to produce the word "rabbit" if asked to produce a word fitting the pattern R_B__T Priming effects are distinct from conscious awareness because they are an implicit memory
Recency Effect
When presented a list of items auditorially, memory for the last 1 or 2 items on the list were more likely to be correct than for items in the middle of the list Stable across a range of conditions (short / long lists, words / nonsense material, items presented slowly or quickly, whether the subject is drunk or sober) Intervening items do affect the curve (delayed recall task); the improved recall for items at the end of the list is lost Unaffected by (even though they improve overall performance): 1) Using common words instead of unusual words 2) Presenting items slowly instead of quickly 3) Presenting concrete words (things that we can visualize/ touch) instead of abstract words Recency effect disappears as a result of delaying the recall attempt with an intervening task Thought to occur because the most recently presented items are still in STM
Role of Familiarity in Chunking
When we "chunk" things, we chunk it into things that are familiar to us Example: 1492177618121941 is a sequence that is too long for STM but can be rearranged into 1492 1776 1812 1941 into important years in American history
Phonological Loop & the Word Length Effect
Word length effect occurs because verbal rehearsal occurs in real time (it takes just as long to rehearse the word "tuberculosis" as it does to actually say it) Longer items take longer to verbally rehearse Longer rehearsal time allows more time for the decay of a memory trace Baddeley believed that people can remember about as many words as they can say in 2 seconds; 2 seconds is thus presumed to be the time limit of the phonological store Subsequent research also suggests that the ability to recall words is also affected by the time it takes to recall the items (not just to rehearse them); it takes longer to recall longer words from memory, contributing to our decreased ability to recall longer words