working memory Flashcards
what was HM’s issues
- Severe amnesia – inability to form new long-term memories for events and facts
- Preserved short term memory
- Preserved procedural memory
what was KFs issues
- Reduced digit span (short term memory - approx 2 items)
* Preserved long-term memory Double Dissociation between long term memory and short term memory
who came up w short term memory model
atkinson and shiffrin 1968
is STm a unitary store
yes for atkinson & Shiffrin
what did Baddeley adn Hitch do
examined effects of concurrent digit load on reasoning, comprehension and learning tasks
– e.g. “A is not preceded by B —— AB”
– Subjects were able to do all these things with a concurrent memory load (albeit with reduced capacity)
• Argues against the idea of a unitary short term store
what is – Articulatory suppression effect
- Subvocal rehearsal impairs memory for words but has no effect on performance of a visuospatial task e.g. chess
- In contrast, chess performance is drastically impaired by performance of a concurrent visuospatial task
- Suggests the existence of two separate systems for temporary storage of information, one phonological and one visuospatial
what is – The Phonological similarity effect
- Phonologically similar words (e.g. ‘fan’ and ‘ban’ are harder to remember than phonologically dissimilar words (e.g. ‘fan’ and ‘cot’)
- Effect doesn’t occur for semantically similar words
- Suggests a phonological code not a semantic code used for temporary storage of words
what is – The Word length effect
- Immediate memory span for short words is greater than long words
- Again suggests a phonological coding system used for short term storage of words
explain Baddeley and Hitch (1974): Working Memory
Working’ component of working memory
• “The most important but least understood component of working memory” (Baddeley, 2003)
• Coordination of resources, attentional control, processing and manipulation of stored information
And the basis of human intelligence
what did Spearman find in 1904
- In 1904, Spearman published a paper examining correlations in children between different disparate measures - academic ability (ratings from teachers, performance in exams etc.) and sensory discrimination
- Found that correlations were all positive
- Correlation between sensory ability and academic ability was almost perfect
- Factor analysis revealed an underlying factor common to performance of many different kinds of tasks (‘g’ factor)
what factors underly the ‘g’ factor
- Gf fluid intelligence (reasoning, problem solving)
- Gc (crystalised intelligence (general knowledge)
what did Kyllonen & Christal give subjects
Kyllonen & Christal gave subjects reasoning tests, such as these…
joe is a farmer. joe is pretty. all farmers are pretty
what is fluid intelligence
problem solving
what is crystallised intelligence
acquired knowledsge
working memory correlates with…
Working memory correlates with fluid, not crystallised intelligence.
Evidence that the ‘working’ component of WM correlates with fluid intelligence
Kane & Engle (2002)
- Two tasks differentially load on the ‘working’ component of WM
- Complex span requires retention of information in WM and in addition requires the active processing, manipulation and updating of this information
- Simple span only requires retention of information in WM
test of fluid intelligence
Ravens Matrices
what is a ravens matrices
subjects see groups of patterns and have to find the pattern that fits into the space
what are is a Structural Equation Model
The structural equation model identified 3 factors underlying performance of these tasks – essentially 3 independent cognitive processes that are differentially engaged by these tasks- working memory (WM), short term memory (STM) and processing speed.
They then examined the extent to which these factors correlated with their measure of fluid (or general) intelligence (gF). As you can see, the correlation between working memory and gF was almost perfect. But the correlation between STM and gF was negative. And the correlation between speed and gF was low.
Working’ component of working memory predicts …
fluid intelligence
fluid, or general, intelligence involves the….
executive attention
what is the Rule Working Memory Task
John Duncan reasoned that there might be a more general process involved in the kinds of WM tasks that correlate so closely with intelligence.
In the complex WM tasks subjects have to remember a set of rules for task performance. He reasoned that it may be this process of remembering and implementing a set of rules that is most closely associated with intelligence.
He gave subjects a task involving a complex set of rules and also the kinds of complex span tasks we saw in the previous studies.
Also gave subjects complex WM span tasks: Solve each maths problem and say word aloud – then recall all 3 words at the end
what did Duncan et al 2012 find
Duncan found that the strongest correlation was between rule WM and intelligence. Other types of WM, such as complex span (digit, spatial or operation) correlated with fluid intelligence but not as strongly.
Thus, it appears that whilst the process of manipulating or processing information in WM might be a key component of fluid intelligence, it is really the construction and use of a set of task rules that underlies individual differences in fluid intelligence.
what did Duncan 2013 argue
- Duncan (2013): Effective fluid intelligence involves construction of a ‘mental program’ for task performance – subdivide goals into sub-goals to break down complex problems into manageable chunks.
- Argues that fluid intelligence involves construction of a mental program for task performance. – subdividing tasks into manageable chunks.
what was Duncan’s 2008 experiment
In this task participants are required to watch letters and numbers appearing in the centre of the screen and report all the letters from one side. Then, when they saw an arrow, they should switch sides and report letters from the other side.
The graph shows the proportion of trials on which subjects made a ‘side error score’ in that they carried on reporting letters from the same side even though they had been cued to switch. As you can see, for the majority of subjects this goal neglect was very rare. However, for a subset of subjects, they neglected roughly 50% of all cues.
As you can see, all of those subjects had IQ scores at the lower end of the scale, between 80 and 100.
What’s interesting about the results is that the subjects all reported knowing what they were supposed to do.
Dancan 2008 results
The graph on the left shows the number of side errors in subjects given partial instructions and the graph on the right shows errors in subjects given full instructions.
As you can see, goal neglect was generally much higher in subjects given full instructions. Indeed, increasing the overall task complexity had the effect of reducing performance in high IQ subjects to the level of that normally achieved in low IQ subjects.
This suggests that it is the overall complexity of the mental program required to complete a task that is important when considering the link between WM and fluid intelligence.
what underlies human intelligence
working memory
working memory is the ability…
• It is the ability to deal with more complex sets of task rules that most closely predicts general (fluid) intelligence