Week 1 Flashcards
Types of Long-Term Memory
Explicit (Conscious):
- Episodic
- Semantic
Implicit (Non-Conscious)
- Procedural
- Priming
- Conditioning
Procedural Memory
Skill Memory: Memory for actions
perform procedures without being consciously aware of how to do them
People who cannot form new LTMs can still learn new skills (e.g., HM)
Mirror Drawing
Figure that you have to trace while looking into a mirror so everything is flipped
as it is done over and over again, you get better at it
Same thing as the next day, the third day better
skill acquisition
Donald Hebb
Canada’s Best Known Psychologist
The Organization of Behaviour (1949)
Father of neuropsychology and neural networks
interested in all aspects of
learning
Priming
Presentation of priming stimulus change’s a person’s response to a test stimulus
Repetition Priming:
Test stimulus the same or similar to the priming stimulus
Called implicit memory because procedures usually
discourage trying to remember the initial stimulus
Performance is often similar whether the person
remembers or not the original presentation of priming stimuli
Priming in Everyday Experience (Perfect and Askew 1994)
Incidental group examined magazine articles to determine what factors make them appealing and readable
There were also 25 adverts (to be ignored) facing the leading page of each of the articles
After asked to rate 50 adverts (25 new + 25 old) and indicate which ones they recognized
Incidental group only recognized 11 per cent of the old ads
Gave higher ratings to old ads on appeal, eye-
catchingness, distinctiveness, and memorability
Cont.
Even though they didn’t consciously remember the previous exposure to the ad, this exposure increased their liking to the ad
Rating changed accordingly based on new ad vs. old ad
Previous exposure nonetheless was encoded into LTM and effected interpretation again when the ad was seen again
Priming in Everyday Experience
Similar to Propaganda effect: more likely to rate statements read or heard before as being true
Even when initially told it was false or didn’t believe it
Involves implicit memory because it can occur when
people are not aware of previously seeing or hearing statement
More likely to be true if it was heard repeatedly
Trump analogy: said something over and over again, expecting it to be true
Classical Conditioning
Pavlov - food (salvation) with dog paired with bell
Dog salivated to the ringing of the bell eventually
Feeling anxious while driving on certain road as previously was given a ticket on that road
Operant Conditioning
Things that are considered rewards that increase behavior
Punishments decrease the likelihood
Reward could also be removal of negative experiences
E.g., people with animals getting shocked and behavior stops when shocked, behavior is changed as animal doesn’t want shock
Skinner box: rat in box with lever and lever dropped food, associated lever with food
Pigeons got food every 15 seconds and they started being weird because they assumed that their actions gave them food even though this wasn’t true
Superstitions
May be used to cope with anxiety
Students kept head down when prof was talking and when prof went to one side of room, students put head up
Prof started to lecture at one side of room eventually
This things unconciously effect our behavior
Explicit Long-Term Memories
Endell tolbin make the first distinction
Episodic is like mental time travel, you can essentially relive that memory
Semantic is thinking about general facts, there is no need to go back to any situation to know these details, they just come to you
Capacity, essentially infinite (estimation is 10 billion facts)
Billions of neurons that can be connected in many ways
Characterizing Memories
Transfer: How is information copied into the store?
Capacity: How much information can the store hold?
Forgetting: How does information get lost from the store?
Representation: What is the format of information in the store?
Retrieval: How is information recovered from the store?
Rehearsal
Rehearsal: a set of techniques/strategies for encoding information into LTM
2 types of rehearsal:
Maintenences: Tries to maintain info for short period of time (phone number); Recycling; keeps info alive in WM
Elaboration: the goal is to remember by repeating to self, thinking of ways to remember it; promotes info to LTM
Hellyer (1962)
Used the Peterson and Peterson (1959) paradigm
The twist was that participants were allowed to repeat the trigram to themselves either 1, 2, 4, or 8 times before starting the distraction task.
More repetitions displayed an increase in performance for the memory test
At 8 times, distraction test basically had no effect
Distraction test wanted to wipe out LTM and STM
This experiment displayed that STM was no longer in the question and LTM is not effected by this test
Recall happens when things get into LTM
Rundus (1971)
Gave people lists of words anc encouraged people to say out loud what they were doing and majority of people did rehearsals
Taped these sessions
Asked them to recall the list of items
Primacy portion involved first words were remembered
Middle items were the most forgotten
Last items were recalled quite well
Looked at relationship between rehearsals and performance
High number of rehearsals = high accuracy of items
remembered (primacy and arthroscopy)
Did not carry at the end
Recency is involved with STM
Sometimes 6th item was rehearsed more than first item
Found that things only rehearsed once were low in ability to be recalled
Linear trend, higher rehearsals = higher correctance
Craig & Lockhart (1972)
U of T
1972: Gus Craik & Bob Lockhart proposed a processing framework for memory rather than simply a store
highly influential view of “levels of processing”
Idea here was that memory may be more than just a storage
We may doing more things with memory
Things to do with stimuli
Levels of Processing
Memory depends on how information is encoded
Depth of processing
Shallow processing:
- little attention to meaning
- focus in physical features
- poor memory
Deep Processing
- close attention to meaning
- better memory
Craik & Tulving (1975)
Got participants to do various types of trials
Participants had to answer questions that came on screen (upper or lowercase font)
Other trials asked if stimulus rhymed with question
Argued that first question was based on structural features
Rhyme required phonemic
Sentence one needed semantic level (deeper level)
Results
regardless performance increased when you went from structural to semantic
Argued that this is strong evidence that depth of processing stimulus influences how well you can remember it
Criticism: time taken for response changes, strutural was quicker than semantic
Exposure of stimulus was important aspect
Macleod (1976)
Is it time spent processing the item?
Had bilinguals decide if word was English/French vs. living/non-living (4 groups)
It took .92, .82, .92 and 1.42 s to make the respective decisions
Accuracy on subsequent (surprise) memory test was 37%, 38%, 48%, and 48%
Time doesn’t seem to be the important factor, but rather it seems to be the depth of processing
Macleod Cont.
Tested if it is time spent processing item or depth of item the important thing
This was a low level of processing to identify english or french
Depth of living or not would be a higher depth prophecy
Time was just less than a second
Living decision was the same as english or french in one action
Surprise memory test: semantic memory was at 48% and english was at 37%
The depth was the more important thing
Criticisms of Levels
circularity = there is no independent measure of depth in the framework (Nelson, 1977; Baddeley, 1978)
context effects – sometimes produce opposite the predicted results (e.g., the transfer-appropriate processing to be discussed)
Largely people looked at performance and assumed that one conditions as deeper in the level of processing
Depends on test that is being done, will effect results
Will not be perfect
Value of Levels
Places emphasis on processes
Introduced a technique - incidental learning - for studying encoding processes
Moved emphasis from passive to memory depends on the processing that is performed on particular test
Introduced a new form of memory test - test of memory was more hidden from observer, incidental learning
Engaging in daily activities and randomly having to remember something
Sometimes you are required to remember that material
Idea is more you do with stimulus, the better you remember it
Better remembrance when item is generated to person (told to remember the world “apple”)
E.g., people who took notes by hand compared to notes typed on computer
Laptop people almost transcribed lecture, hand people are not quick enough so they have to try to summarize
Argued that people on laptop are transcribing and keeping info at low level of processing and hand people are trying to summarize it and are thinking of it at a semantic level
Representation
Jenkins & Russel (1952); Bousfield (1953)
Random words to be learnt by free recall
People recalled recent items first
Clusters occurred (e.g., chairs, couches, tables)
Argued that there is some kind of organizational principle which tends to be semantic
Things were piled in a room and participant went to every pile and tried rememberin things from there
Kintsch & Bushchke (1969)
Looked at peoples performance and where they made errors
Things that were semantically related to another
When people made errors at the end, they were sound alike errors
End of list are stored in STM which is predominantly encoded acoustically so this makes sense; errors were due to WM
Beginning errors were items were rearrangements of semantics so it was a semantic confusion (LTM, auditory)
Argued that LTM is encoded in meaning and WM/STM is encoded in sound
Nelson & Rothbart (1972)
24 “paired associates” learned until perfection, forget for a 4-week retention interval, return for relearning forgotten pairs
Identical, Control, Acoustic
We don’t try to remember things in exact situation but the jist
Meaning associated with it
LTM evidence is not only semantics but there is always roon for acoustical
They forgot a bunch of pairs after a month
The researchers created a new list
Either paried it again or repaired with a new item or repaired with something similar to acoustics but it was different item
Something was still there with old experience that allowed them to learn this again (ebbinghuas notion of savings)
Results
Found performance was low for control items
Identical: performance went up 70%
Acoustic: depends if acoustic info was somewhat in LTM, performance was at about 50%
Suggests that even though meaning is stored in LTM, acoustic may be stored as well
Kolers (1975)
People showed advantage of reading same material in same orientation
Savings were held even a year later
Benefit to the original stuff they read before
Visual representation allowed them to remember better when they experienced it again
Visual info is there in LTM but it is not the predominant
Semantic is general, but there is some visual and auditory
Interference
Is interference responsible for loss in memory
Jenkins & Dallenbach (1924)
Participants required to remember list of items and were made to go to sleep or stay awake
If RI is critical to LTM, then sleeping immediately after learning should help memory
Learnt ebbinghuas type nonsense syllables
Sleep gave benefits in remembering information
People did worse staying awake
But longer sleep still worsened performance
Results
Retroactive interference plays a role
Sleep helped reduce information after knowing the letters
Yaroush, Sullivan, Ekstrand (1971)
No REM sleep meant no dreaming so no extra information displayed
REM sleep is crucial in memory retention
Less things that interfere with memory, the better the performance
Benton Underwood
Developer with Leo Postman of the interference theory of forgetting
emphasized role of proactive interference (PI) in LTM
Theory looked at how interference affected memory - proactive interference
Compared his studies to 1800s Ebbinghaus and his results were way worse
BUT ebbinhaus was his patient and he learnt a lot more with himself
Proactive Interference
Repetition made it harder to remember the new list
More lists exposed to, the worse performance had gotten
Some evidence for pro and retroactive interference
Loss of Information from LTM
Both RI and PI seem to play a role
Material in LTM but for some reason, cannot be retrieved
This is called retrieval failure
Some argue there may be a role for decay
Retrieval from LTM
Vast amounts of info to search
At 1 ms per memory and 10 billion memories, it would take 58 days to retrieve a given memory
LTM may be “content addressable” to optimize access and speed
Cannot access memory in parallel
This is like a serial like search
We have some kind of organize system which clusters things together
LTM has a system which allows to have a retrieval cue which leads to the access of a certain part of LTM
Context Dependency
Emphasizes the match between encoding (study) and retrieval (test)
* Encoding Specificity
* State-Dependent Learning
* Transfer-Appropriate Processing
Degree to which the way you study and test and the strength of the match
Endel Tulving
U of T
best known memory researcher in the world
* some of his key ideas:
* semantic/episodic
* subjective organization
* encoding specificity
Organization doesn’t always have to be based on semantics
E.g., saying cat after table, you will be able to remember it even if list was shuffled (subjective organization)
Tulving & Thomson (1973)
Encoding Specificity Principle: the idea that the way in which information is encoded determines the optimal way to retrieve that information
* The better the match the more likely you will be able to retrieve it
You want to make the test and study as similar as possible
Godden & Baddeley (1975)
Had 4 groups of deep sea divers that learned a list of words:
* Group 1: Studied on Land/Tested on Land
* Group 2: Studied on Land/Tested 20 feet underwater
* Group 3: Studied 20 feet underwater/ Tested 20 feet underwater
* Group 4: Studied 20 feet underwater/ Tested on Land
Results
land group did better than underwater group in testing
Argued that environment and stress levels impacted this perhaps
Underwater group that tested underwater did better than studied underwater and tested on land
The other two groups performed around the same
This proved that people did better when their settings were as similar as possible
E.g., studying in quiet environment and testing in quiet environment would help in better grades
Match environments that you are in
Eich & Metcalfe (1989)
State-Depending Learning: Learning is associated with a particular internal state
* Better memory if person’s mood at encoding matches mood during retrieval
Looked at internal state instead and how it influenced performance
Happy situations or sad situations
Filled questionnaire to confirm mood and then were told to study a list of words
When test state matched study state, they did better
Sad and sad and happy and happy did better than other groups
Morris et al. (1977)
Transfer-appropriate processing:
- Memory task results improve if the type of processing used during encoding is the same as the type during retrieval
Results
People engaged in rhyming test
When told to encode, one group of participants were asked to answer questions about the rhymes
Both groups were put into rhyming test
People were better when their encoding was similar to their retrieval (rhyming encoding and rhyming test)
This task, keeps it at acoustic level so the meaning encoding should be the better performance task
This suggests that sometimes this may not be the case
You will learn better if you space out studying
You learn things more while producing the results (told to remember fruit with letter a… apple)
Bransford & Johnson (1973)
Title or picture provided context to help understand the passage
Title or picture only helped if given before the passage
Participants required to try to remember a written passage
Context given ahead of time could really help but only if it is given ahead of time
This helps you organize the information and put it in its place
Izawa (1970)
Is it better to study more?
80% for 3 tests and 90% for 1 test 5 study
Karpicke & Roediger (2008)
study-and-test experiment with word pairs
Group 1: studied and tested all words/sessions (81%)
Group 2: studied only words missed in the previous test; tested on all words (81%)
Group 3: studied all words; tested only on words missed in previous tests (36%)
Factors that Lead to Better Memory
- Levels of Processing (Craik & Lockhart, 1972)
- Distributed Practice (Ebbinghaus, 1885)
- Generation effect
- Organizing to-be-remembered information
- Retrieval practice
Other Factors
Make images that incorporate things help people do better
Imagery group did better for boat-tree
Being a group can be beneficial
Helping answer other questions and being able to relate to others
