WM Flashcards
What’s a possible function of the PHONOLOGICAL LOOP in Baddeley and Hitch (1974) WM model?
learn new vocabulary –> Baddeley, Papagano, and Vallar (1988): patient PV capacity to learn pairs of meaningful words was unimpaired, but she was not able to learn foreign language vocabulary.
What’s a possible function of the VISUO‐SPATIAL SKETCHPAD in Baddeley and Hitch (1974) WM model?
imagery –> involves maintaining an image which can then be mentally manipulated.
Shepard and Feng (1972): paper folding problems. Average time to decide whether the arrows on the cubes would match as a function of number of imaginary folds necessary to reach that decision.
Provide a definition of CENTRAL EXECUTIVE, according to Baddeley and Hitch (1974) WM model.
A supervisory system that controls other systems such as perceptual, memory and response systems. Linked to distinction between automatic and controlled behaviour (latter requires EF) : PFC.
Describe Cowan’s (1988) EMBEDDED‐PROCESSES MODEL
WM content is determined by the focus of attention.
A central executive controls focus of attention, which acts on recently activated features from LTM.
Capacity limit is the result of a limited‐capacity focus of attention (= approx. 4 objects / chunks at the same time).
Some information in LTM can be active even when it is not in the focus of attention. This information is more easily retrieved than non‐activated portions of LTM. There is decay of activation when the activated information is not in the focus of attention.
e.g. the Sperling task: stimulus presentation activates letter representations in LTM. This activation rapidly decays. But, can be retained longer if brought into the focus of attention.
The UNITARY MEMORY MODEL is an alternative to STM/LTM distinction, what’s its basic assumption?
A single item is attended, the rest needs to be retrieved from memory (Nairne, 2002; McElree, 2006…)
Recency effect is not a demonstration of STM!
Describe some evidence for the Unitary Memory Model.
Wickelgren et al. (1980): examined retrieval speed as a function of serial position for lists of 16 sequentially presented items. Results: Strong recency effect for multiple objects, but only the final item had a reduced RT because only the final item of a list was attended. Oztekin et al. (2010) Replicate with brain imaging. Task: which of two items had been presented in study list? Results revealed the typical recency effect. RTs reduced as a function of recency. The active set = items presumed to reside in WM according to the standard model. Hippocampal activation (reflects LTM) was higher for the active set than for the passive set, and it significantly reduced for the final item (focus of attention). Conclusion: Passive and active set are both in LTM. But the last item is not retrieved from LTM.
How is possible to solve the conflict between results suggesting only 1 item is attended (the rest in LTM) and those of vSTM studies (Luck and Vogel, 1997; Vogel et al., 2001): 3 to 4 items held in a visual short-term store.
Oztekin: serial presentation ≠ vSTM studies: simultaneous presentation
Possible solutions:
Only one chunk is attended, but with simultaneous presentations 3 or 4 items can be chunked together.
With simultaneous presentations rapid shifts of attention can rehearse up to 3 or 4 items, so that only one item is attended at any given time.
Multiple items can be attended, but serial presentations may shift attention to the newest object.
Oberauer and Hein (2012): distinction between broad (capacity 4) & narrow focus of attention (cap. 1).
What is the capacity limit in Working Memory? Provide evidence.
Cowan et al. (1999): Subjects matched pictures (primary task) and heard a spoken list of digits (secondary memory task). On a few infrequent trials a memory response screen appeared and subjects attempted to recall the digits. Around 4 items of unattended speech recalled.
Enumeration reaction times: Subjects are required to indicate the number of objects briefly displayed. Reaction times hardly increase up to 3 or 4 items, but increase more steeply for higher numbers.
Multiple object tracking: up to 3 or 4 is relatively easy, but performance declines more rapidly with more objects (e.g. Yantis, 1992).
According to Goldman‐Rakic’s WM Model [based mainly on primate studies (e.g. single‐cell recordings)] what is the role of dlPFC and vlPFC in WM?
dlPFC = WM for spatial material (‘where’) & vlPFC = WM for objects (‘what’)
Neural activity in dlPFC: cue square is flashed at a particular position, causing the neuron to respond. The square goes off , but the neuron continues to respond during the delay. The fixation X goes off, and the monkey demonstrates its memory for the location by moving its eyes to where the square was.
According to Petrides’ Theory of WM, what is the role od dlPFC and vlPFC in WM?
Assumes division of PFC into at least two separate processes: maintenance (vlPFC) and manipulation (dlPFC). Storage or information is in the posterior cortex.
OBERAUER & HEIN (2012) – THREE‐EMBEDDED‐COMPONENT MODEL OF WORKING MEMORY
Name and describe the three components.
- Activated part of LTM: holds information relevant to task available.
- Region of direct access/broad focus of attention: 4 items, binds them into new structures*=Cowan’s focus of attention.
- Single item focus of attention: selects one item as target of next cognitive operation.
OBERAUER & HEIN (2012) – THREE‐EMBEDDED‐COMPONENT MODEL OF WORKING MEMORY
describe the Two-cue, two-probe recognition paradigm. For wich of the three components does it provide evidence for?
General procedure:
a. participants learn two STM lists
b. then one of the lists is cued (=relevant)
c. Recognition task (two times): Probe word (P1) > is it in the cued list? Second list cue – Probe word (P2) > is it in the cued list?
Critical manipulation: variable cue‐probe interval. Rationale: STM lists need to be held in the broad focus (provides candidate set of items for direct access). When a cue is presented the irrelevant list can be removed from the broad focus, but this takes time: at short cue‐probe intervals the irrelevant list is still in the broad focus, whereas with long intervals it is removed from broad focus; but: not forgotten!
Other information can be outsourced to activated LTM, and quickly retrieved back: makes it possible to flexibly select subsets of required information.
OBERAUER & HEIN (2012) – THREE‐EMBEDDED‐COMPONENT MODEL OF WORKING MEMORY
Describe some evidence for the narrow focus of attention.
Recognition–retrieval dynamics: last item of a studied list is accessible at the highest rate = (only) the last item is already in the focus of attention (no retrieval required)
Object-switch costs: performance costs resulting from switching between items in WM –> last object is still in the focus of attention (no retrieval required)
Retrieval benefits from cueing individual items: cue helps in change-detection tasks, compared to no/invalid cue, but: only one cue; two cues have no benefit –> only one item can be focused at a time.
Limitation of visual search to a single search template: in visual search tasks people use one item in WM as a template of the target. If search for two targets simultaneously: accuracy decreases.
Sternberg’s (1966) memory search/scanning task.
Data suggest serial scan of STM items (size effects). [in the scanner: dlPFC]
OBERAUER & HEIN (2012) – THREE‐EMBEDDED‐COMPONENT MODEL OF WORKING MEMORY
Sternberg’s (1966) memory search/scanning task. Retrieval from STM
e.g. Memory set: A E L F C J; Probe (present in set Y/N?): J?
What’s the result and how does Oberauer interpret it?
Oberauer’s interpretation: narrow focus of attention shifts within the broad focus (region of direct access).
Testing the model:
Sternberg’s set-size effect used to examine the contents of the region of direct access: more items in the broad focus means greater RT (more shifts of the narrow focus within the broad focus)
Examine the set-size effects of both the relevant list and the irrelevant list!
OBERAUER & HEIN (2012) – THREE‐EMBEDDED‐COMPONENT MODEL OF WORKING MEMORY
Why isn’t one single focus “zoom-lens” (Cowan) more likely than the two foci of attention?
This would not permit both at the same time Oberauer (2002): two lists of digits, first cuing one list as relevant (broad focus), then mathematical operations of list objects, observing object-switch costs (narrow focus).
