Chp 5 - Working/Short Term Memory Flashcards
Atkinson and Shiffrin’s (1968) Modal Model (5)
Modal model of memory: The standard model of memory consists of 3 primary components
The sensory registers
Short-term store
Long-term store
- Control processes
- Encoding
Sperling 1960 (STM recall task)
Full report condition:
* Participants recalled on average 3-5 individual letters
Partial report condition:
* When cued to remember a specific row, participants recalled 3-4 items correctly
concluded that a short-lived sensory memory registers all or most of the information that hits our visual receptors, but that this information decays within less than a second.
Controlled processes (proposed by Atkinson & Shiffrin)
the part of the standard model of memory responsible for the active manipulation of information in STM
Dynamic processes associated with the structural features that can be controlled by the person and may differ from one task to another.
An example of a control process that operates on short-term memory is ‘rehearsal’. Other examples are strategies used to help make a stimulus more memorable and strategies of attention that help you focus on information that is particularly important or interesting.
Baddeley’s Model of Working Memory (4)
This model assumes that each component has a limited capacity and is relatively, not entirely, independent of the others.
1) the phonological loop
2) the visuo-spatial sketchpad
3) the central executive.
4) episodic buffer.
Phonological similarity effect
Information can be represented:
* Wickelgren (1965)
* Participants tend to confuse similar-sounding letters (e.g. B and
C) but not similar-looking letters (e.g. O and Q)
Brandimonte et al. (1992) (pipe cup thing)
demonstrated that materials can be processed in visual form if the verbal channel via the phonological loop is made inaccessible.
visual processing in working memory
Interfere rehearsal
- Those who had to repeat “lalala” were not afriad to attached a label to the objects,
- Less to think of a pipe as a pipe
- Easier to think of the removed part from pipe, as a cup
Identified the shape easier due to no verbal label
Not repeated:
- Rehearsed the pipe as a pipe, labelled
support Baddeley’s theory
Brooks (1968) (mechanisms)
When task and respond rely on same mechanism, response is worse
- Spatia-spatial maxes out capacity of the visuospatial sketchpad (vs. spatial verbal)
- Verbal-verbal maxes phonological loop
- Tax out same resources
- Incorporation of 2 things
The sensory registers
(Modal Model) (3)
- E.g., iconic and echoic memory
- High capacity but limited duration
- an initial stage that holds all incoming information for seconds or fractions of a second
Short-term store
(Modal Memory) (5)
- Information can be represented:
- Visually
- Phonologically
- Semantically
- Duration: very short (~20 seconds)
- Capacity: 7 +/- 2?
Long Term Store
(Modal Memory)
- can hold a large amount of information for years or even decades
1) the phonological loop
(Baddeley) (3)
- processes sounds and is responsible for speech based information.
- this includes sounds that are processed in one’s mind.
- For example, the phonological loop is used in learning new vocabulary, problem-solving, math problems, and remembering instruction
2) the visuo-spatial sketchpad
(Baddeley) (3)
- responsible for processing visual and spatial information.
- It can be fed either directly, through perception, or indirectly, through a visual image.
- The visuo-spatial sketchpad allows people to store images of objects and their locations.
3) the central executive.
(Baddeley)
- incorporates information from the phonological loop, the visuo-spatial sketchpad, the episodic buffer, and from long-term memory
- the switching of retrieval plans, time sharing in multitasking, selective attention, suppressing irrelevant information, daydreaming, and temporary activation of long-term memory.
4) episodic buffer.
(Baddeley)
temporary store that integrates information from the other components and maintains a sense of time, so that events occur in a continuing sequence.
- seen as a place to temporarily integrate information gathered from the phonological loop, visuo-spatial sketchpad, and long-term memory.
- controlled by the central executive, yet it transfers information into and out of the long term store.
Encoding
(Modal Model)
the act of taking in information and converting it to a usable mental form
o i.e., encode auditory info into sensory memory, if the info is transferred to STM then it is said to have been encoded to STM
Brown and Peterson: causes of forgetting
forgetting might be caused by the passage of time before testing (decay)
Brown-Peterson Task
- short-term memory task showing forgetting caused by proactive interference
- People were given a 3-letter trigram and a 3-digit number and asked to count backward by 3s
- This distractor task requires lots of attention and prevented the rehearsal of information (backward counting)
Results
- With an increasing time span less and less information remains in short-term memory
- Waugh and Norman thought the distractor task might be a source of interference
- This study suggested that forgetting was influenced by number of intervening items, not just the passage of time
Proactive Interference (PI)
Keppel and Underwood (1962)
Older material interferes forward in time with recollection of current stimulus
e.g., stimuli learnt from an earlier list cause difficult in recalling later words
Retroactive Interferences (RI)
Keppel and Underwood (1962)
Newer material interferes backward in time with recollection of older items
e.g., recalling items from list 1 but recalling from list 2
Release from PI (2)
Wicken’s study showed Release from PI
Proved that release from PI occurs when the decline in performance caused by proactive interference is reversed because of a switch in the to be remember stimuli
- Performance deteriorates because of the buildup of proactive interference
Release from PI also occurs when change is semantic
(e.g. rmb flowers to rmb animals)
- The more related items on the fourth list were to the original category, the less release from PI experienced
- Short term memory uses semantic memory
Serial Position Curves
It is a graph of item-by-item accuracy on a recall task.
The y axis is the probability that a participant will remember a given item.
The x axis refers to the items by their original position in the study list.
Recall tasks
- Free Recall
- Serial Recall
Free Recall
People can recall items from memory in any order
Serial Recall
People must recall the items in their original order they were presented
- More Difficult of the two tasks, often requiring rehearsal
Primacy effect
The elevated recall of the first items seen
Recency effect
The elevated recall of the last items seen
Glanzer and Cunitz (1966): Manipulating Serial Positioning Curve
- Had groups of participants wait different periods of time before recalling the items
- Found that individuals who had a counting task (10-30 seconds) before recalling items showed very low recency, but the primacy recall was unaffected
- Shows that the recency portion is susceptible to interference, but the primacy portion was not impacted by interference and was more permanently stored in long term memory
Engle’s Controlled Attention Model
Kane & Engle, 2002
- Working memory is knowledge from long-term-memory that is currently being attended to
- Irrelevant information is suppressed or removed from working memory
Attention and Working Memory According to Engle’s Controlled Attention Model
According to this view, working memory is similar to a mental spotlight
- This model places emphasis on a person’s ability to deliberately retrieve and focus on information and place it in the spotlight
2 Aspects to Attention (Engle)
There are two important aspects of attention control:
1. The scope of attention: how much information can be attended to
2. Control of attention: how good people are at directing what is and is not attended to
Dual Task Method + purposes (3)
Performance is examined by having a person do a secondary task, one that consumes working memory resources, at the same time as some primary task
Purposes:
- Complete Dependence
- Intermediary Dependence
- Complete Independence
Example of Dual Task Method: Brooks (1968)
F task
Participants asked to mentally trace F border and then to either verbally indicate or spatially indicate “yes” or “no” upon reaching every corner if “this corner is at extreme top of figure”
Classify every word as a noun or not and then verbally or spatially indicate “yes” or “no” it is or is not a noun
Logie, Zucco and Baddeley (1990) (squares task)
Overall task
- People saw a grid of filled in squares then they disappeared, they had to point at the area where the squares were using their visual-spatial memory
- There were two groups of secondary tasks;
- mental addition task (irrelevant to the visual-spatial processing) or
- an imaging task (irrelevant to phonological loop)
Secondary task with visual span task
- Visual span performance dropped 15%, mental addition only disrupted visual memory modestly
- When secondary task was visual imagery; visual memory span dropped about 55%, the visuo-spatial sketch pad was stretched beyond limits (spatial-spatial)
Secondary task letter span task
- Mental addition was very disruptive, leading to a 65% decline, same pool of resources
- Imaging task depressed letter span scores only a modest 20%, different part of working memory
Other work
Dual tasks impact encoding, rather than retention of information in working memory
Key Concepts of Dual Task Method (4)
Emphasizes the limited capacity of working memory
Provides evidence for Baddeley’s model of working memory
Simultaneous processing of two separate stimuli w/o interference supports phonological loop & visuo-spatial sketchpad concepts.
Dual tasks congruent in WM processing demands demonstrate interference (retroactive) mechanisms in action
Dual Task Method Relation to Baddeley’s Model of Working Memory
When one node of working memory is overloaded…
1. Task performance suffers
2. Begin to pull resources from the central executive to compensate
Conway , Cohen and Bunting (2001) - examined working memory span and its relation to the classic cocktail party phenomenon
Findings: about 65% of people with low memory spans detected their name in the dichotic memory, compared to 20% in people with high memory spans.
Low memory span – difficulty blocking out or inhibiting attention to distracting information.
Kane and Engle (2003) – used the Stroop task
- No difference in stroop effect high and low when words were always mismatched or when half were presented that way.
- 20% of words mismatching – low span made twice as many errors as high span
Due to rarity, low span seemed less able to maintain task goal in working memory
Sanchez and Wiley (2006) – tested people with different memory spans, giving texts that included illustrations
Illustrations irrelevant to the text – related to topic, but not support
Performance better if working memory capacity were to focus on relevant details
Low Span = more likely to be “seduced” by the irrelevant details in pictures – and were more difficulty controlling contents of their current stream of thought
1st Experiment:
Rosen and Engle (1997)
(name animals)
administered a verbal fluency task to high-span and low-span people, had to list animals as quickly as possible for up to 15 minutes
Findings: High-span individuals did better than low-span individuals with differences seen as early as 1 minute in
2nd Experiment:
Rosen and Engle (1997)
fluency test and a dual-task
- groups were tested in fluency test and a dual-task, were shown digits that showed up and had to hit a key every time 3 odd numbers appeared in sequence
Findings: Performance was lower for the animal naming task for high-span individuals only, low-span individuals showed no decrease in performance
Rosen and Engle Conclusions (2)
- High-span people could produce harder animal names using a controlled attentional process using working memory
- Digit monitoring task also used working memory resources up causing the high-span group to perform like the low-span group when performing the dual task
Kane and Engle (2000) and Brown-Peterson Task Proactive Interference (PI) (3)
people with low-span working memory experience more proactive interference (search a wider range of information, leaves them vulnerable to the interference of irrelevant information)
High Span people used their controlled attentional processes to combat PI when they had a secondary simultaneous task.
Low Span people search for a wider range of knowledge, making them more prone to having irrelevant information intrude on retrieval
Moore, Clarke, and Kane (2008) and Runaway Trolley Car
Working memory capacity as a mediator for decision-making
- More consistent reasoning for people with higher working memory capacity
Copeland and Radvansky (mem and logic)
- Memory span and logic problems
- People with high memory span used more complex ways to solve the logic problems and solved more (syllogisms)
- Can hold more memory in active working memory
What are the models of memory learnt?
- Atkinson’s and Shiffrin’s Modal Model
- Baddeley’s Model of Working Memory
- Engle’s Controlled Attention Model
