Memory (Dr Caleb Owens)

Question 1

Part of explicit memory, what is autonoetic awareness? (L6)

Answer 1

Autonoetic awareness is awareness of the fact that I remember myself to be there.

Question 2

In what circumstances would an iconic memory become a short term memory before it decays?

Answer 2

When I pay attention to it, an iconic memory can become a ST memory before it decays.

Question 3

Define iconic and echoic memories. What types of memories are they?

Answer 3

Iconic and echoic memories are sensory memories.

Iconic memories are pictorial representations and echoic memories are auditory. They both decay rapidly

Question 4

Using the traditional multi-store model, what type of code is used to store ST compared to LT memories?

Answer 4

ST memories are in phonological code. If rehearsed enough to become LT, they then become semantic, meaning they are stored in code of their meaning.

Question 5

In the traditional memory store model, what’s the difference in rate of forgetting between ST and LT memory?

Answer 5

ST memory decays rapidly (after 30s). LT memories are only forgotten due to interference rather than decay.

Question 6

Give an example of STM’s sensitivity to semantic encoding which is evidence against the traditional multi-store memory model.

What does this evidence imply about the relationship between STM & LTM?

Answer 6

STM encoding capacity is usually 7+/-2, however, we can remember a sentence of 16 words easily. This is because it has meaning; this demonstrates we’re using our LTM simultaneously to draw knowledge of this meaning.

This evidence implies a bi-directional relationship, as opposed to the uni-directional relationship inferred by the muli-store model.

Question 7

What is proactive interference, and how are we released from this?

How does this demonstrate that LTM is involved during encoding STMs?

Answer 7

Proactive interference takes place when we are overloaded by too many items in a similar category (eg fruits). We are released from this when we change semantic categories (eg fruits - flowers).

Proactive interference demonstrates that when encoding new information, LTM is drawn upon for semantic category information.

Question 8

STM is sensitive to levels of processing. What is the evidence supporting this, and how does it disprove the multi-store model?

Answer 8

People encode more effectively when asked to contextualise what they are encoding. For example ‘you find these in a city’ (trucks) is then recognised as a word already seen more frequently than ‘How many letters does this word have’? Or ‘Does it rhyme with treat’?

This implies that when attaching semantic categories we encode more successfully than sounds (rhyming) or print (counting the letters). This again adds evidence ot the interactive (bidirectional) relationship between STM & LTM.

Question 9

Describe and explain the three components of Baddeley’s original (1974) and revised (2000) models of working memory, addressing how the revised model explained integrated memory traces in LTM.

Answer 9

Baddeley’s original model of working memory involves the central executive with 2 slaves; a phonological and a visuospatial loop. The central executive coordinates these 2 slaves, which is why it’s possible to add load without compromising WM performance when the information is from different modalities (eg we can listen to music while working on a maths equation, but can’t easily do a maths equation whilst someone is counting aloud). However, this model was not seamless because it doesn’t explain how things progress to LTM.

In 2000 he revised the model by adding an ‘episodic buffer’ which integrates information from both modalities - phonological and visuospatial, in order to create a unified trace for LTM.

Question 10

Using the Wisconsin card sort task, how is a person with frontal lobe damage from a car accident likely to perform in Friedman and Miyake’s (2012) WM test which uses switching?

Answer 10

When the task type is switched part way (eg sort cards by colour, now sort by shapes. Rule is made known by participant being told ‘yes that’s right’ or ‘no that’s wrong’), a person with frontal lobe damage will find it very hard to switch tasks midway.

Question 11

Friedman and Miyake’s (2012) updating, shifting and inhibition tasks are designed to load your ______ function.

Briefly describe each task.

Answer 11

Friedman and Miyake’s (2012) updating, shifting and inhibition tasks are designed to load your executive function.

Updating; remember the last 3 letters (must throw out oldest and update with 2nd two oldest and new number each time).

Shifting; task switches mid-trial (match shape to match colour)

Inhibition; inhibit saccade to flash on right of screen and report viewing arrow on left

Question 12

When factor analysis was applied to Miyake’s data, which EF factors did Friedman identify?

Answer 12

There was a common EF (executive function) factor across all 3 updating, shifting and inhibition tasks, and a seperate factor for both updating and shifting but not for inhibition.

Question 13

What did Ranganath et al (2003) find about encoding, recognition, WM & LTM using face recognition trials during fMRI?

Answer 13

Ranganath et al (2003) found no difference in brain regions active during WM (face recognition) and LTM (delayed face recognition) tasks. They did find a difference between brain regions used during the encoding and recognition phases. More brain regions were active overall during the recognition phase, encoding was more concentrated around the visual cortex.

Question 14

What does the encoding specificity principle (ESP) describe about recall, and how does this differ for recognition?

Answer 14

The encoding specificity principle describes the phenomenon that when a memory is encoded, the surrounding environment (context) can act as a retrieval cue when the memory is recalled. For this reason memories encoded underwater will be more successfully retrieved whilst underwater.

However, a recognition task is unaffected by context, since the recognised item is itself the retrieval cue.

Question 15

What’s the main difference between the encoding specificity principle and transfer appropriate processing?

Answer 15

The main difference between the two is that transfer appropriate processing focusses more on the process involved while encoding memory, whereas the ESP focuses on the environmental circumstances present during encoding. Keep in mind that ESP later came to include emotional state whilst encoding.

Question 16

In Squire’s 1987 cognitive science of LTM memory systems, which 2 memory types are defined as explicit (declarative) and which 3 are defined as implicit (procedural- nb this is one!)?

Answer 16

Declarative, or explicit memory, encompasses episodic memory (one’s own experiences, includes autobiographical) and semantic memory (facts, general knowledge).

Procedural, or implicit memory, encompasses procedural memories (motor skills, habits & tacit rules), classical conditioning effects and priming (implicit activation of concepts in LTM).

Question 17

Classical conditioning effects (eg conditioned emotional reactions), priming (implicit activation of concepts in LTM) and procedural memory are all part of which memory type?

Answer 17

They are all part of implicit (non-declarative, unconscious) procedural memory.

Question 18

Episodic memory including autobiographical memory and semantic memory (facts and general knowledge) are both part of which memory type?

Answer 18

Episodic and semantic memory are both part of declarative, conscious explicit memory.

Question 19

How did transfer appropriate processing challenge the notion of depth of encoding (where deepest encoding produces memories more successfully). Give an example.

Answer 19

Transfer appropriate processing showed that matching the process of encoding with the type of retrieval had a stronger effect than depth of encoding.

For example, using rhyming to encode, which is weaker than semantic encoding, is actually more successful when the retrieval task cue uses rhyme eg ‘did it rhyme with regal’? as opposed to standard implicit recognition (have you seen this word before)?

In this example, encoding conditions were rhyming (does it rhyme with legal)? and semantic (is it a bird)? where the item is ‘eagle’.

Question 20

Estimates of STM storage were originally ‘the magic number’ 7 +/-2. As memory theories updated towards the concept of WM which utilised LTM, what is the new ‘pure’ capacity estimate and what other cognitive process does this align with, and how is that demonstrated?

Answer 20

The updated estimate of WMC is 3/4 chunks (max 5). This aligns with attentional ability, demonstrated through the MOT (multiple object tracking tasks) where you have to keep your attention on the moving dots.

Question 21

What memory type are recall and recognition tasks testing? Justify.

Answer 21

Recall and recognition tasks are testing explicit, declarative memory because subjects know their memories are being tested and are asking to declare what they’ve seen.

Question 22

Fragment completion tasks are tests of what type of memory? Explain how these tests work.

Answer 22

Fragment completion tasks are tests of implicit memory, because participants are unaware their memory is being tested.

Participants are shown lists of words and then, after some time (even a different day, as took place in our lectures) participants are asked to complete words which have missing letters and could be multiple possible words. If participants complete word with word previously seen, the word with missing letters has effectively acted as a cue for the word previously seen, therefore a successful memory trace was created.

Question 23

Which memories have more longevity, explicit or implicit?

Answer 23

Implicit memories have more longevity than explicit memories.

Question 24

What is the evidence that HM’s bilateral hippocampal lesion (medial temporal lobe) provided about implicit and explicit memory systems?

Describe the experiment which evidenced this.

Answer 24

When HM’s hippocampus was lesioned bilaterally, he lost retrograde memory up until about 2 years before the lesions, and also anterograde memory explicitly. However, he was able to improve incrementally at the challenging mirror drawing task, which shows that his implicit anterograde memory was still intact (procedural memory; skills learning).

Question 25

Does depth of encoding effect explicit and implicit memory differently?

How about emotional manipulation?

Answer 25

Yes, depth of encoding doesn’t effect implicit memory whereas it has a robust effect on explicit memory.

Emotional processing also effected explicit memory retrieval but not implicit.

Question 26

What did Warrington and Weiskrantz (1970) find when they used free recall, recognition, word fragment identification and word stem completion tasks to compare people with amnesia with people without amnesia? Identify which tests are explicit and which are implicit.

Answer 26

Warrignton and Weiskrantz (1970) found that people with amnesia had impaired explicit memory formation (free recall and recognition tests), whereas their implicit memory formation was not effected (word fragment identification and word stem completion), compared to participants without amnesia.

Question 27

Match between modality at encoding effects implicit but not explicit memory, what is meant by this? Give an example which happened in class.

Answer 27

Modality at encoding for implicit memory could be the font used when presented a list of words. When the implicit memory test, a fragment (word) completion task is given, whether the same font is used or not will effect the outcome.

This is not the case for explicit memory; we don’t need to font for our study notes to match that in the exam.

Question 28

Match between modality at encoding effects implicit but not explicit memory. What did Weldon and Roediger (1987) find when they used pictures and words as modalitites?

Answer 28

Weldon and Roediger (1987) found that modality at encoding (pictures vs words) effected implicit memory (word fragment completion test) but not explicit memory (free recall task). In their results, people who encoded pictures and completed the word fragmentation task had significantly compromised retrieval compared to people who encoded words and then completed the word fragmentation task.

On the other hand, people in the explicit memory condition of free recall performed as expected in that memories encoded as pictures were retrieved more frequently than those encoded as words. Since free recall is a verbal/ written test, this shows that the modality did not effect retrieval negatively as it did in the implicit word fragmentation test.

Question 29

• Amnesia: affects explicit but not implicit memory performance
• Depth of encoding affects explicit but not implicit memory task
• Match between modality at encoding and retrieval affects implicit but not explicit

These all examples of single dissociations. Why, and why do they not prove that there are seperate ‘systems’ for explicit and implicit memories?

Answer 29

These are examples of single dissociations, because they show a difference between explicit and implicit retrieval.

Question 30

Critique of the explicit/ implicit memory systems concept includes Roediger (1990), who claimed that transfer appropriate processing explains why different processes are used to both encode and retrieve memory traces. Explain his theory, demonstrating how memory tasks are either perceptually and conceptually driven.

Answer 30

Roediger (1990) explained that implicit memory tasks are perceptually driven whilst explicit memory tasks are conceptually driven.

For example, encoding for an implicit memory task by reading words merely activates LTM, and the retrieval process is familiarity; hence both processes are perceptually based.

On the other hand, encoding for explicit memory tests is elaborate; participants engage with conceptually-driven strategies and recollection involves deliberate intentional retrieval. Therefore both encoding and retrieval are conceptually driven.

Question 31

Explain the remember / know distinction and how it adds further evidence to transfer appropriate encoding. How did Konstantinou and Gardiner (2005) demonstrate this?

Answer 31

Konstantinou and Gardiner (2005) tested levels of processing by having participants encode faces in shallow (how round is this face?) VS deep (what is this person famous for?) conditions.

During retrieval, if participants claimed they remembered a face, they were asked follow-up questions such as ‘do you remember seeing this face before’? or ‘does this face look familiar but you’re not sure where you saw it’?

They found that faces encoded in the deep condition were remembered more successfully, but there was no difference in ‘know’ (familiarity) between deep or shallow encoding.

This adds further evidence to the transfer-appropriate processing because it shows that shallow and deep encoding have the same familiarity outcome, whereas deep encoding has more successful retrieval (‘remember’) than shallow encoding.

Question 32

What does remember/ know suggest about an overall view of the memory system, and what is the neuroscientific evidence produced by Wheeler & Buckner (2004) to back this up?

Why does this make sense given my understanding of the function of the cortical areas involved?

Answer 32

The remember/ know procedure suggests an overall memory system which doesn’t have separate compartments for explicit and implicit memories, but instead one system which interacts with information such as depth of encoding appropriately for the type of retrieval needed (remember vs know / recollection vs familiarity).

Wheeler and Buckner (2004) used neuroimaging to show that both procedures use the same parietal regions, but that the ‘remember’ / recollection process also uses the hippocampal medial temporal area.

This makes sense when thinking about parietal areas as association areas (‘this looks familiar’) and the hippocampal medial temporal area as the memory zone active when retrieving recently encoded memories.

Question 33

Describe Jacoby and Kelly’s (1992) process dissociation method, and how did it separate recollection and familiarity (remember from know) using inclusion and exclusion tasks?

Answer 33

The process dissociation method used 2 stem completion tasks; the inclusion task and the exclusion task.

The inclusion task includes measures both recollection and familiarity. Participants are instructed to complete word stems with words they remembered seeing at encoding, or otherwise just to use any word they could think of.

The exclusion task measures familiarity only. Participants are asked to complete the stems with words they had not seen before. By asking participants to use only words they had not seen before, if they used a word which was included at encoding, it shows that the word was familiar but they did not recollect it.

Question 34

Jacoby and Kelley (1992) used process dissociation and tested the effects of dividing participants attention during encoding for recollection and familiarity.

What did they find?

Why were their testing measures for dissociation ‘pure’?

Answer 34

By using the dissociation method, Jacoby and Kelley (1992) showed that dividing participants attention affected recollection but not familiarity.

The exclusion condition of the process dissociation method is quite a pure way of separating familiarity from recollection, because participants are asked to complete the word fragments only with words they haven’t seen before.

Question 35

Which of these is involved in the ‘know’ procedure: Guessing, familiarity or recollection?

Answer 35

Familiarity is involved with the ‘know’ procedure. You can’t guess knowing, you either feel you know it or you don’t.

Question 36

In the memory systems view, procedural memories, priming, classical conditioning & habituation & sensitisation are all parts of which system?

Answer 36

Procedural memory, priming, classical conditioning and habituation and sensitisation are all part of non-declarative memory.

Question 37

In the memory systems view, episodic and semantic memory are part of which system?

Answer 37

Episodic and semantic memory are part of declarative memory.

Question 38

What is the evidence which supports the memory systems view? Name 2 types of dissociation.

Answer 38

There are 2 dissociation types which are evidence for the memory systems view:

1. Amnesiacs suffer declarative episodic memory loss but not non-declarative memory loss. This can be viewed as explicit memory loss but not implicit. eg HM can learn mirror drawing
2. explicit memory is effected by depth of encoding but implicit memory isn’t.

Question 39

What is the generation effect in the context of our RR?

Answer 39

The generation effect demonstrates that it’s more effective to generate a word, for example ‘the alcohol produced when grapes ferment is calle ____’, than when the word is simply read in sentence.

Question 40

In the context of our RR, what is the difference between generating and producing a word? Describe the tasks.

Answer 40

In the ‘generate’ word task, I saw an incomplete sentence such as “fermented grapes produce _____”.

Then I was prompted with an option such as ‘wine’ or ‘alcohol’, and I responded yes or no depending which I had generated.

In the production task I had to type my response, either wine or alcohol.

Therefore generating can be seen as ‘thinking of something’ whereas producing is either typing, writing or reading aloud.

Question 41

What executive processes are needed for N-back tasks, on top of coordination of demands? Think of the 3 designed by Freidman and Miyake (2012).

Answer 41

Where an N-back task involves presenting a series of letter in different locations, updating, shifting and inhibition are required.

Updating information held in WM, inhibiting unnecessary information (letters/ locations which are only 2 back if the task is a 3 back), and shifting between the location attended and the letter attended.

Question 42

Is working memory training transferable to other tasks? Consider the evidence put forward in week 2s tutorial, Jaeggi et al (2008) and Dunning, Holmes and Gathercole’s (2013) study of at-risk children, which had 3 conditions; no intervention, non-adaptive and adaptive training.

Answer 42

Jaeggi et al (2008) claimed that WM training was transferable to fluid intelligence; but they only used one measure (Ravens) to test Gf and their control group was passive, meaning that they received no training at all. An active control group who received some similar exercises that did not train WM would have provided more convincing evidence.

The At-risk children study found that trained children only showed improved WM; it was not transferrable. The adaptive group showed more improvement than the non-adaptive group.

Question 43

What did Melby-Lervag, Redick and Hulme’s (2016) meta-analysis show about transfer affects of WM training?

Answer 43

The meta-analysis showed that transfer effects were more prominent when the transfer test was most similar to the WM training.