NHB (Neurobiology and Human Behaviour) Lecture
Notes: Memory
Cognitive Neuroscience of Memory
Memory is needed for learning
Measures of memory (these are all only correlates)
● Conscious (behaviour)
○ Objective:
◆ Accuracy of recall
◆ Reaction time of recall
○ Subjective
◆ Remember/know (do they remember/know)
◆ Confidence (how confident are they on that memory)
◆ Qualitative ratings (from a memory how much can you
remember of the sight, smell, sounds)
● Unconscious
○ Priming (will respond quicker (behaviour) even if you don’t think
you’ve seen word before)
○ Conditioning
○ Habits (riding bike (muscle memory))
● Psychophysiological
○ Skin galvanisation, HR
● Electrophysiological
○ ECG, EEG, EMG, MEG
● Haemodynamic
○ PET, SPECT, fMRI
Double dissociation
● Changing parameter A effects response A but not B
● Changing parameter B effects response B but not A
● So responses must be distinct
Stages
● Encoding information into memory
● Storage of information in memory
● Retrieval of information from memory
Early ideas of memory
● Primary memory (psychological present, consciousness, RAM) -> STM
● Secondary memory (psychological past, not in consciousness) -> LTM
Multi-store model
● Sensory store —(attention)—> STM —(rehearsal)—> LTM
○ Forgotten by decay, displacement, interference respectively
● Criticisms
, ○ Memory is not strictly serial, can have LTM without STM
◆ Patient KF: TPJ lesion, impaired STM (digit span) but not LTM
(primacy effect)
◆ fMRI also shows inferior frontal and parietal cortex activity
for STM
○ STM and LTM are not unitary and uniform
◆ Patient KF: worse STM for auditory stimuli (letters and digits)
than visual stimuli
◆ So possibly different STMs for different kinds of information,
like sensory stores
Sensory store, one for each modality
● Experiment: iconic store
○ Flash grid of letters for 50ms
○ People were only able to recall for ~1s
○ If prompted immediately (by light) to recall a row instead of the
whole grid, they had higher accuracy since they were able to recall
the fewer letters before the 1s ran out
● Experiment: echoic store
○ Dichotic listening
○ Only noticed messages in both ears were the same if played within
~2s of each other
STM
● Can hold 5-7 items
○ Enhanced by chunking and rehearsal (vocally or sub vocally)
(longer an item is held in STM by rehearsal, greater likelihood of
LTM storage)
● Often unreliable, distraction (interference) usually causes forgetting
● Experiments
○ Digit span
◆ Can recall 7±2 digits, in English (other languages take longer
to say digits)
◆ Same for letters and words, so its done by items (chunks) not
length
◆ People can choose to chunk things differently
○ List words
◆ Remember list of 20 words, rehearse out loud (so rehearsal
can be measured)
◆ Serial position curve of word position against accuracy
◆ Primacy effect from LTM due to rehearsal
◆ More rehearsal of word, higher accuracy of recall
◆ Recency effect from STM
◆ If you stop rehearsal after presentation for a while before
recall, it stops recency
Notes: Memory
Cognitive Neuroscience of Memory
Memory is needed for learning
Measures of memory (these are all only correlates)
● Conscious (behaviour)
○ Objective:
◆ Accuracy of recall
◆ Reaction time of recall
○ Subjective
◆ Remember/know (do they remember/know)
◆ Confidence (how confident are they on that memory)
◆ Qualitative ratings (from a memory how much can you
remember of the sight, smell, sounds)
● Unconscious
○ Priming (will respond quicker (behaviour) even if you don’t think
you’ve seen word before)
○ Conditioning
○ Habits (riding bike (muscle memory))
● Psychophysiological
○ Skin galvanisation, HR
● Electrophysiological
○ ECG, EEG, EMG, MEG
● Haemodynamic
○ PET, SPECT, fMRI
Double dissociation
● Changing parameter A effects response A but not B
● Changing parameter B effects response B but not A
● So responses must be distinct
Stages
● Encoding information into memory
● Storage of information in memory
● Retrieval of information from memory
Early ideas of memory
● Primary memory (psychological present, consciousness, RAM) -> STM
● Secondary memory (psychological past, not in consciousness) -> LTM
Multi-store model
● Sensory store —(attention)—> STM —(rehearsal)—> LTM
○ Forgotten by decay, displacement, interference respectively
● Criticisms
, ○ Memory is not strictly serial, can have LTM without STM
◆ Patient KF: TPJ lesion, impaired STM (digit span) but not LTM
(primacy effect)
◆ fMRI also shows inferior frontal and parietal cortex activity
for STM
○ STM and LTM are not unitary and uniform
◆ Patient KF: worse STM for auditory stimuli (letters and digits)
than visual stimuli
◆ So possibly different STMs for different kinds of information,
like sensory stores
Sensory store, one for each modality
● Experiment: iconic store
○ Flash grid of letters for 50ms
○ People were only able to recall for ~1s
○ If prompted immediately (by light) to recall a row instead of the
whole grid, they had higher accuracy since they were able to recall
the fewer letters before the 1s ran out
● Experiment: echoic store
○ Dichotic listening
○ Only noticed messages in both ears were the same if played within
~2s of each other
STM
● Can hold 5-7 items
○ Enhanced by chunking and rehearsal (vocally or sub vocally)
(longer an item is held in STM by rehearsal, greater likelihood of
LTM storage)
● Often unreliable, distraction (interference) usually causes forgetting
● Experiments
○ Digit span
◆ Can recall 7±2 digits, in English (other languages take longer
to say digits)
◆ Same for letters and words, so its done by items (chunks) not
length
◆ People can choose to chunk things differently
○ List words
◆ Remember list of 20 words, rehearse out loud (so rehearsal
can be measured)
◆ Serial position curve of word position against accuracy
◆ Primacy effect from LTM due to rehearsal
◆ More rehearsal of word, higher accuracy of recall
◆ Recency effect from STM
◆ If you stop rehearsal after presentation for a while before
recall, it stops recency
