Cognition and Neuroscience
28 September 2020 13:53
Upload Date: Monday 12:00 noon. Assessments:
Online Q&A sessions: Monday 16:00 Unseen Essay (summer exam) - 50%
Weeks 4,7 & 9 1500 word essay - 30% =
CAQ's - 20% - 230/320 = 71%
Topics: ▪ Uploaded on Mondays; Remains open until week 5/11
1. Episodic Memory
2. Prospective Memory Lecturers:
3. Semantic Memory - Dr Jayne Freeman: j.e.freeman@reading.ac.uk
4. Cognitive Ageing - Dr Dan Jones: dan.jones@reading.ac.uk
7. Executive Processes
8. Autobiographical Memory
9. Embodied Cognition
10. Impaired Cognition
PY2CNS Page 1
,-------------------
17 May 2021 10:56
PY2CNS Page 2
,Episodic Memory
28 September 2020 14:28
LTM:
Declarative memory: Memory you can express
• Episodic: events or 'episodes'.
• Semantic: facts and concepts. 'General knowledge'
Procedural memory - no conscious awareness. E.g. procedures. How? Riding a bike.
• Implicit - events influence our decisions and actions without our awareness.
Examples in psychology:
Remembering a list of words
- Includes contextual information such as when and where. Who was there, what happened after. Rich in contextual
information. Unique to EM
○ Autonoetic awareness is also unique. It's the self-awareness of the situation, feeling like you can relive the part
through memories.
Processes:
- Encoding: acquiring information:
○ Attention, mood, intention to learn, etc. all influence encoding. Most important factor is the type of processing that
we carry-out during encoding while processing the event.
▪ That is the key idea contained in the LoP (levels of processing) theory.
- LoP: Craik & Lockhart (1972)
○ During recall, we don’t recall an exact copy if the event but rather how it was carried out and processed.
○ They argue that we can process at different levels of depth.
○ Continuum: From shallow, surface level processing to deeper processing.
○ Deeper processing leads to better recall.
- LoP: Craik & Tulving (1975) experiment Example: the word holiday
○ Ps read a list of unrelated words (bat, coffee, shield, coin), and used incidental (not told that there will be a test so Shallow processing:
learning is not intended) learning. Focus on the perceptual or physical features such as how it
○ Gave people an unexpected recognition test. is written: font, colour, etc.
▪ Given a list of the original list of words mixed with new words. Had to report if they have seen the words before. Deep processing:
○ Conditions were used to manipulate the types of processing used when encoding. Focus on the semantic features of the word - the meaning
▪ Shallow: Decide if the word was written in upper or lower case (bat or BAT) of the word: If the word is pleasant, what are the
▪ Intermediate: Rhyme judgement (Does the word rhyme with hat) associations, synonyms, etc.
▪ Deep: Meaning of the word. Fill in the blank in sentence to see if it makes sense.
□ Later Ps were shown the word and asked if they’ve seen it before or not.
Significant increase in the amount of words that were recognised.
Shows that processing at a deeper level does lead to better recall later. Well established finding.
○ Problems:
▪ Don’t have a good independent way of measuring depth. How do we know that making a judgement on
meaning leads to better processing that judging its appearance. Hard to know the varying levels of processing
that go on. This creates a circular argument.
▪ Interpretation: when we're seeing a benefit we assume because it involved in deeper processing but could just
be that deeper processing may be more effort. We need to put in more effort in understanding the wording
instead of just seeing what it looks like. Could be that the effort is what causes LoP effect and not the depth.
▪ The benefit of the semantic processing that we see depends on the kind of tests that you use to examine
memory.
□ Morris, Bransford and Franks (1977) study
Used 2 types of encoding;
1. Semantic type: Shown a word and generate a word that associates with it (Hat - scarf)
2. Phonological encoding: Generate a word that rhymed with the word on the list (hat - mat)
Also manipulated what types of test were used
1. Standard: Shown old and new words and asked to select the old.
2. Rhyme: Ps had to say whether the word on the test list rhymed with a word that they had
seen (Test list: cat, answer yes, as it rhymes with hat (old list))
□ Found: Semantic encoding paired with standard testing showed a greater recall later on than phonological
based test.
However: Found the exact opposite when looking at Rhyme recognition when paired with
phonological encoding
= Performance is best when the type of processing used at encoding, matches the processing
required at test
◊ Transfer-appropriate processing - processing depends of type of test.
- What underlies the LoP effect
○ Craik & Tulving (1975): Semantic may lead to better processing as it produces more elaborate (making connections
between the words and other stored knowledge) encoding of the information of the words on the list.
▪ Elaboration hypothesis:
▪ Key concept: We are more likely to remember something if we can make connections between the item and
other things that we already know.
▪ Argue that semantic processing will always lead to more elaboration and connection to stored knowledge than
non-semantic coding
▪ Support: Craik & Tulving (1975):
□ Those in semantic (deep) condition were asked if the words made sense at the end of a sentence.
□ This condition showed higher LoP effect as well as found a Congruity Effect.
People are more likely to remember the words that made sense at the end of a sentence rather
than those that didn’t make sense (congruent over incongruent).
▪ Congruity effect:
□ When the word fits at the end, the sentence helps create more elaboration for the word. Sentence
provides connect and other things that are associated with the word.
□ Semantic processing is needed for both congruent and incongruent sentences, but the congruent
sentences provide even more elaboration = congruity effect.
○ Hunt & Elliot (1980): Semantic lead to more distinctive encoding - Lets those items stand out more.
▪ How well we remember something later on depends on the distinctiveness of the information encoding: how
well the item stands out in memory from other things.
PY2CNS Page 3
, well the item stands out in memory from other things.
▪ Evidence: Bransford at al. (1979):
□ Gave Ps sentences to read:
'A mosquito is like a doctor because both draw blood' Second is more elaborate and provides more info, but the first is
'A mosquito is like a racoon because they both have a head, legs and jaws'. more distinctive.
□ Found: Ps more likely to remember the first sentence.
This tells us that if something is unusual in some way, were more likely to remember it later on than
if it is not distinctive. (Better LoP).
▪ Argues that semantic processing (meaning of the word) leads to better memory because it becomes distinctive.
□ items are more likely to be distinct or unusual in their meaning as opposed to how they look.
▪ Eysenck (1979): used a non-semantic phonological (sounds) task. Asked Ps to pronounce those words in unusual
ways. Still leads to a distinctive memory trace as its unusual to - i.e.- pronounce 'glove' as 'cove'
□ Found: memory was just as good following this phonological task as it was following a semantic processing
task.
□ Tells us that we can get distinctiveness from different types of processing: semantic and non-semantic, we
would just usually gain more distinctiveness from semantic processing thus seeing the typical LoP effect
○ Winograd (1981):
▪ Compared the role of distinctiveness and elaboration in people's ability to recognise faces (more naturalistic).
□ Ps were assigned to different conditions: 1) Ps were shown a series of faces and asked to spend time
scanning each face (encoding lots of info like features - elaboration) and then rate the most distinctive
feature (distinctiveness). 2) Ps shown a series of faces: told what was distinctive about face and instructed
to only focus on that feature alone (just distinctiveness)
□ Tested which faces people had seen before and found no difference in memory performance. This
suggests that distinctiveness is the most important factor in how well a memory trace is made.
This doesn’t mean however, that elaboration has no role to play. As the more info we ha ve, the
more likely we are to detect a distinctive feature.
Bower et al. (1969):
- Organisational material is remembered better - the extent to which the items are connected together.
- Ps were presented with 112 word list in which words were either presented in an organised hierarchy or just randomly
○ Given 4 trials to learn the words: after each presentation they were asked to recall the words
○ Found that the organised hierarchy condition were great at recording the words: after the 4th trial, 100% of the words
were recalled.
○ Randomised hierarchy had good recall at 70% but it was still significantly less than organised.
Generation Effect:
- Glisky & Rabinowitz (1989):
○ We're more likely to remember something If the info is generated on our own rather than heard or seen.
○ Study:
▪ Manipulates what people did at encoding, and what they did at test.
▪ Study phase: condition 1) people presented with a word and asked to read it
2) Ps were presented with a word fragment (some letters missing), and were then asked to
generate a word using the remaining letters
▪ Test Phase: Condition 1) Read the word and decide if they recognise it from the study phase (Read + Recognise)
2) Ps given fragments, asked to generate a word using the fragments, and then had to
report if they have seen the word before during the study phase
○ Found:
1) People more likely to remember words if they had generated them for themselves during the study phase than
if they had just read them. (Standard Generation Effect)
2) Memory was better if people generated the words at both study and test than only generating at study and
reading at test.
3) Best recognition performance was found when ps were asked to generate the words during study and test
phases. (transfer appropriate processing):
-lep----t at study and -lep----t at test better than…
-lep----t at study and e--p-a-t at test
Self-Reference Effect:
- Rogers, Kuiper & Kirker (1977):
○ The finding that: were more likely to remember things if they are relevant to us or if we've thought about them in
terms of their relevance.
▪ The personal relevance of info has a powerful effect on memory
○ Study:
▪ Presented Ps with a list of adjectives (i.e. kind, tall, etc.) and asked them to use one of 3 different types of
encoding to complete tasks:
1) Ps were given a non-semantic task : decide whether the words rhymed with another word.
2) Semantic encoding condition: decide if the words meant the same as something else
3) Self-relevance condition: Ps have to decide how descriptive a word was of themselves
▪ Looked at people's ability to recall the adjectives
○ Found:
▪ Recall was significantly better for those in the self-relevance condition than in the either of the other two
conditions. Would have found LoP effect if the second condition was more significant but instead they found
that the best memory of all came from the Ps who rated the words in relevance to themselves
□ Thus, Self-relevance leads to greater processing through allocation of more attention
Retrieval: Retrieve or remember the information.
- Why does it fail?
- Storage failure: Decay
○ The older the memory trace the more likely it is to be forgotten (Thorndike, 1914)
- Rather than just decay, it's more likely that the reason as to why we fail to store information is that other events occur an d
other info comes along that then interferes with the original memory of the event.
Interference: New info overwrites/replaces older info
- Jenkins & Dallembach (1924):
○ Asked Ps to learn a list of nonsense syllables and then to either sleep or stay awake for the same period of time.
Found that when syllable recall was tested those who slept had much better memory for syllables.
PY2CNS Page 4