1
Q
But consider some other forms of learning that are informally said to rely on “__________ memory”. Explain
A
“muscle”
- Tasks like riding a bike, tying ur shoelaces, or texting on a cellphone
- Each involves a motor skill that u eventually perform easily & on “auto-pilot” @ least with enough practice
2
Q
Memory also explains the impressive decision making ability of __________. Explain
A
experts
- Ex, when a radiologist examines a chest x-ray & detects a faint abnormality that others have overlooked, this pattern recognition skill is driven by prior training & experience.
- Ex, experienced firefighter describes an intuitive sense of danger that’s difficult to verbalize, leading him to call his crew out of a seemingly
stable building shortly before it collapses.–> When later pressed, these experts can’t fully explain their impressive behaviour, just that it “comes naturally” to them after so much first-hand experience.
3
Q
anterograde amnesia
A
- the loss of ability to form new memories
–> like in movie memento, Leonard Shelby has anterograde amnesia –> Like a forensic detective who has just arrived @ the scene, he continually searches for clues to help make sense of his current situation - A similar form of anterograde amnesia has been studied using the fascinating case of “Patient H.M.”
4
Q
Deese-Roediger-McDermott paradigm
A
- aka DRM Paradigm
GOOGLE
- a psychological experimental method for studying false memories, where participants are shown a list of semantically related words (e.g., web, insect, bug, fright, fly, arachnid, crawl, tarantula) & then incorrectly recall or recognize a related but unpresented word (Ex, spider)
–> This is probably cuz all the similar words that u saw, like web, insect, small, creepy, and so on, led u to reconstruct ur experience in a way that would fit with spider also being presented. –> Put simply, ur brain might have falsely inferred that spider was there given all the other similar info that u encoded.
-
5
Q
Human memory is more ___________________ in nature.
A
reconstructive
- u might naturally think of memory as a sort of recording, like filing a document in a cabinet, filming a video on a camera, or saving a file on a computer. –> But u would be WRONG
–> every time u rmr some event or piece of info, it not necessarily being retrieved as it was originally stored. –> Instead, it is an inference or reconstruction of ur prior experiences.
6
Q
Encoding
A
- describes when info is first being learned.
- How info initially enters into memory. –> A selective process that is highly dependent on attention.
–> ex, by focusing on someone’s face to discern whether they r friend or foe, u encoded whatever it was u paid attention to.
7
Q
Storage
A
- describes what info is being stored in the brain & how
- Concerns how the record of memory is maintained over time.
–> This record is not fixed–> remains flexible & can be modified over time. –> Ex, after telling the same story many times, u may adopt & maintain inaccuracies or fabrications without being aware of it.
8
Q
Retrieval
A
- describes how we access prior experiences to make use of them in the present.
- The act of recovering stored information.
–> Memory retrieval is dependent on retrieval cues= a key piece of info that has the potential to activate a memory in full. –> Ex, smell of lilacs, a strong cue, might make u think of picking flowers @ grandma’s farm
9
Q
Encoding vs storage vs retrieval: New route to school example
A
- imagine that u find a much faster route to school & take it
consistently for a few days.
–> Here we might say u have ENCODED some aspects of the route’s landmarks or its general spatial layout. - then u have to stop taking the new route cuz of a one-month road closure. –> During this 1 month, how r ur prior experiences with the route represented in ur brain? & how does this stored info change over time? –> These r matters of STORAGE
- the route re-opens again, & u try to re-navigate it solely based on what u rmr. –> How
many landmarks do u rmr? & how well do u navigate the route relative to 1 month ago?–> These r matters of RETREVIAL.
10
Q
Testing our hypotheses: how human memory experiments r conducted.
A
- During initial study phase, participant is usually shown a series of simple items (like words or pics) & asked to process them in some way.
–> Ex, a researcher might ask 1 group of participants to generate a bizarre image of each word in their mind & ask a separate control group to
silently read each word - After a set amount of time where the info is stored in memory= the retention interval, participants undergo a test phase where they r asked to rmr something about the items they saw during the study phase.
–> Performance in the test phase thus allows us to examine specific hypotheses about the way memory works. –> Ex, using the bizarre imagery & control groups we just discussed, some theory might hypothesize that bizarre imagery enhances memory, & so more of the og items should be recalled following bizarre imagery than silent reading. - things get a lot more complicated cuz, there r many diff ways u can test someone’s memory.
11
Q
The Retrieval Cue.
A
- is basically any piece of info that can be used to access other info that is stored in memory.
–> Ex, As 2 friends banter back & forth, 1 piece of info triggers retrieval of another, shaping the flow of the convo. –> While Jane tells a funny story about her cute puppy, Yara is reminded of a similar incident when her dog was an even cuter puppy.
–> Like the cue–response process studied by behaviourists, 1 piece of info acts as a cue to trigger retrieval of something else stored in memory. - the same way that a convo is guided by what retrieval cues r provided by ur convo partner, a memory test will reveal diff things depending on what retrieval cues r provided by the experimenter.
12
Q
There r 2 common ways researchers test a participant’s ability to rmr items from an encoding phase: List them
A
- free recall
- recognition
13
Q
free recall test
A
- very few retrieval cues r provided
–> the participant is simply asked to rmr as many items as they can from the encoding phase with no additional info to help them rmr.
14
Q
recognition test
A
- the participant is shown several items: some of which r “old”, meaning they were previously shown in the study phase; & some of which r “new”, meaning they were not.
–> participant is asked whether each of these words is old or new, & the key question is how well they correctly identify the old items as old & the new items as new.
–> in this case, the old items
r being presented as u originally saw them in the encoding phase, & so the retrieval cues provided during this type of test r thought of as very strong.
15
Q
Part of the complexity of memory is that encoding, storage, & retrieval r all highly ___________________. Explain
A
interconnected
- rmr ex with a bizarre imagery and a silent reading group. –> The results come back & show that the bizarre imagery group outperformed the silent reading group. –> Did this occur cuz bizarre imagery led to better encoding of the items, more durable storage of the items, or an improved ability to retrieve the info during the test phase?
- We’ll hold off on the details for now, but memory researchers have found some clever ways to disentangle these diff aspects of memory performance. –> gives u a better a sense of why precise terminology is so important!
16
Q
Question 1: imagine a similar competition where contestants r shown an image of common scene for only 5 seconds—let’s say a messy bedroom—& later asked to rmr as many objects from the scene as they can. What type of memory test
is this?
A
A free recall test –> If a contestant lists a # of objects in the messy bedroom without any reference to the objects (perhaps a list of the objects in the image), then the
contestant is participating in a free recall test
17
Q
Much of what we know about this topic of how info acc gets stored in memory comes from studies on people who…
A
…experience amnesia & other memory deficits cuz of brain damage or other abnormalities
–> In fact, one patient in particular has fundamentally changed the way we think about human memory= patient H.M
18
Q
Patient H.M
A
- 1 of most important case studies in all of psych.
- Plagued by chronic seizures,
27-year-old H.M. underwent a drastic surgery in 1953 to remove a large portion of the hippocampus on both sides of the brain. –> this key brain area is linked to the formation of new memories. - The surgery = success cuz there was a marked reduction in seizures.
–> but there was an unintended consequence: H.M. developed severe anterograde amnesia, meaning he had great difficulty committing new info to memory.
–> Every experience he had from that point forward felt new to him; any convo, any person he met, any new word he encountered—it all quickly faded from his memory shortly after the acc experience. (~20 seconds) –> For the rest of his life he couldn’t tell u where he lived, what his last meal was, or even how old he was, leaving him stranded in a perpetual present. - yet some parts of his memory remained remarkably intact. –> He had a normal IQ (IQ unaffected), could retain info long enough to have a short convo (short term memory intact), & could learn complex new motor skills like drawing images seen through a mirror.
19
Q
why were some forms of memory impaired in H.M. while others remained normal?
A
- Striking dissociations like those displayed by H.M. have led memory researchers to divide memory into various subsystems that each serve a diff role.
–> The multi-store model, 1st proposed by Atkinson & Shiffrin in 1968, is perhaps the most well-known way to divide up memory into subsystems.
20
Q
The Multi-store Model
A
- 1st proposed by Atkinson & Shiffrin in 1968, is perhaps the most well-known way to divide up memory into subsystems.–> This model also yielded the terms short-term & long-term memory which have become part of our everyday language
- assumes that incoming sensory info is 1st stored in a short-term memory buffer
–> short-term memory buffer operates like RAM on a computer: info held temporarily in short-term memory is available and held “online” for the near future, but not stored permanently.
–> However, info held in short-term memory can be transferred to the long-term store for more permanent storage if it is sufficiently rehearsed. –> This model therefore proposes 2 separate memory stores, hence the term “multi-store”.
21
Q
According to the og multi-store model, memory can be divided into 3 separate subsystems: list them
A
- sensory memory
- short-term memory
- long-term memory
22
Q
Sensory Memory
A
- when u scan ur environment, ur eyes dart across the scene, stopping for brief intervals to gather info.
- This behaviour seems similar to taking ‘snapshots’ @ diff locations & yet u perceive the world as a continuous representation.–> This is partly cuz each snapshot is briefly maintained & then replaced by the next. –> Sensory representations r displaced or overwritten by new incoming info from the same sensory modality.
- This transient maintenance of perceptual & physical info from the very recent past= sensory memory
–> Ex, think about using a bright sparkler on a dark night–> As u move the sparkler through the air to write ur name, u perceive the light persisting briefly as letters even tho the sparkler occupies only 1 location at any point in time.
23
Q
Many of ur senses maintain a similar transient representation of an experience via sensory memory. Elaborate
A
- Iconic memory= Sensory memory by our visual system.
- Echoic memory= Sensory memory for our auditory system.
- Haptic memory= Sensory memory for our tactile system.
24
Q
how long do sensory memories persist?
A
- George Sperling tested the longevity of sensory memories by briefly (e.g., 50 ms) presenting arrays of 3 rows of #s simultaneously.
–> Participants viewed each array & listened for a tone of high, medium, or low pitch that indicated which row they should report.–> that were accurate @ reporting the #s when the tone sounded immediately after the array was presented - However, when the tone sounded just 1 second later, they recalled practically nothing, illustrating the extremely fast rate at which sensory memory decays
- Interestingly, the decay rate seems to depend on modality:
–> iconic memory= last up half a second
–> echoic memory= up to 4 seconds
–> haptic memory= up to 2 seconds
25
Short-term Memory
- Info that is selected from sensory memory enters consciousness & is maintained in the short-term or working memory buffer.
- It is thought to operate like RAM on a computer, whereby the selected info is held online for a short period of time but not necessarily stored permanently
26
How long does info remain in short term memory?
- Info is only maintained transiently in short-term memory if it goes unrehearsed. --> Peterson & Peterson (1959) tested participants’ short-term memory for 3 consonants after varying delays (3–18 seconds). --> to prevent them from rehearsing during the delay, they were told to count backwards from a three-digit # by 3s/4s. Participants were good @ recalling the consonants after a 3-second delay, but recall was abysmal after an 18-second delay.
- Through rehearsal info can be maintained in short-term memory for longer.
27
George Miller discovery
- George Miller found that, for most short-term memory tasks like memorizing a random sequence of #s, people can rmr about 7±2 items.
--> Beyond this load the demand on short-term memory becomes strained, so not all the info can be rehearsed, leading some of
the info to quickly fade from the short-term store.
28
Chunking
- While 7±2 items might not seem like much, Miller found that info can be further organized into "chunks" of meaningful packets, expanding the capacity of short-term memory.
- List 1 & List 2 both contain 16 letters, but u will find that List 2 is much easier to maintain in short-term memory. --> The key diff is that the letters in List 2 can be chunked into 5 bits of info that form familiar letter groupings
--> Letter groupings that form words pack even more info into a single chunk.-->Ex, u may only be able to rmr 7 random letters of the alphabet, but u can also rmr 7 diff names of animals that constitute a total of 30 letters
- this way we can overcome the limited capacity of short-term memory, using prior knowledge to impose structure on the info we wish to maintain.
29
Chunking: study
- In 1 classic study, Chase & Simon (1973) asked chess experts & novices to study the placement of chess pieces on a board.
--> When the chessboard had pieces placed @ random, both experts and novices were likely to make several errors when recalling the positions of the pieces. --> but when the display had pieces in plausible positions from acc games, chess experts outperformed the novices & were much better @ recalling the positions.
- The familiar groupings of the
pieces thus acted as meaningful “chunks”, allowing the chess experts to retain more of info in short-term memory than novices.
30
Working Memory
- Working memory is the modern understanding of short-term memory--> researchers still debate whether there is a clear distinction between short-term memory.
- u can loosely conceptualize working memory as an ‘upgrade’ to short-term memory in that it seeks to better explain how people actively manipulate info in short-term memory.
- Consists of 3 short-term buffers (or stores): the phonological or articulatory loop, the visuospatial sketchpad, & the episodic buffer.--> these 3 short-term stores r thought to transiently hold pieces of info so they can be further manipulated
31
Phonological loop
- aka articulatory loop
- Encapsulates the og notion of short-term memory.--> It's a temporary online store that can maintain7±2 bits of phonological info for a short period of time.
--> The phonological loop maintains info that can be rehearsed verbally
--> Ex, u might engage ur phonological loop if u rehearsed a phone # verbally as a way to rmr the series of #s.
32
visuospatial sketchpad
- new addition to the working memory model.--> It is thought to temporarily represent/store & manipulate visual info.
--> Ex, u might engage ur visuospatial sketchpad by evoking a mental map of how to get to work.
33
Episodic buffer
- new addition of the working memory model.
- It's thought to draw on the other buffers (phonological loop & visuospatial sketchpad) as well as on other stored long-term memories, combining these inputs to construct a coherent & multimodal event over a brief period.
--> This aspect of working memory is engaged when rmring specific past episodes.
--> Ex, in thinking about the last time u ordered from ur favourite pizza restaurant, u might recollect what the menu looks like & where it's pinned onto ur fridge.
34
The working memory model also includes a _______________ _______________
central executive
- Coordinates & manipulates the info stored in the working memory buffers.
--> allows for higher-order coordination of info temporarily held in the 3 stores.
- Simply put, this addition of the central executive allows short-term memory to be much more flexible than it was originally conceived to be.
35
Working Memory: diff components summary
- Phonological Loop= temporarily stores auditory info.
- Visuospatial Sketchpad= temporarily stores visual info
- Episodic Buffer= uses input from buffers & long-term memory to retrieve past events & briefly maintain them.
- Central Executive- allows for the coordination & manipulation of info held in the stores.
36
Movie Memento & Patient H.M. : Such extreme cases of anterograde amnesia suggested something puzzling
that short-term memory & long-term memory r supported by separate systems in the brain.
37
Info is not stored in isolation. Explain
- instead, long-term memory seems to be highly organized according to some higher-order structure, especially pertaining to semantics or meaning.
--> Ex, we tend to recall related info in clusters or groups even if it was learned in a random order --> Imagine u 1st studied the following word list: apple, screwdriver, hammer, lemon, pear, level, wrench, & peach.
--> When later asked to recall all the words u r likely to 1st list all the fruits & then all the tools
- So-called “semantic networks” capture this type of organization, using nodes to represent concepts & lines to connect them --> Ex, u might imagine a node for “fruit” & a line connecting it to another node for “vegetables”.
--> Related concepts r thus said to be closer or more interconnected, & activation of a concept is akin to throwing a pebble into a pond such that the activation spreads (or ripples) throughout the network to activate other neighbouring concepts.
38
semantic network: Pizza example
- the concept of “pizza” makes me think about pineapple & the time they mistakenly added anchovies, but it also reminds me that I have not had dinner yet & makes me hungry.
--> Then, when I think of how hungry I am, I naturally think of hungry hippos. --> In this semantic network the concepts of ‘pizza’, ‘toppings’, ‘dinner’, ‘hungry’ & ‘hippos’ r all related concepts that tend to activate 1 another.
39
complex info is also thought to be organized in long-term memory as ___________.
- schemas= With regard to memory, it is the mental frameworks for interpreting the world around us based on prior experiences. --> these knowledge structures guide encoding & retrieval.
--> Ex, u might have a schema for “dinner party”: u usually ring the doorbell, greet the hosts, help prepare food, eat the food over convo, help clean up, lounge around a bit, then head home.
- study by Brewer & Treyens (1981) that showed the power of schemas
--> They had participants wait in an office that was set up like a typical office in the 1980s: some tables, a typewriter, & a coffeemaker.--> also contained some unusual things like a screwdriver & it purposely did not contain some things u'd expect in an office, including books.
--> after 35 secs in room, participants were taken into another room & asked to describe all the items they rmred seeing in the room.
--> nearly 33% reported seeing books whereas only 10% reported seeing the screwdriver that acc was there. --> show us how strongly schemas (or other prior knowledge) shape our memory
40
long-term memory
The permanent storage of our memory.
--> Breaks down into declarative (semantic & episodic) & nondeclarative memories (implicit & procedural).
41
Retrieval from long-term memory
- Simply put, we don't retrieve some discrete entity in long-term memory that represents an event, but rather our brain attempts to reconstruct that event through schemas or other prior knowledge stored in long-term memory.
--> Think to change blindness experiments from Attention chapter where 2 images r presented rapidly in alternating fashion & ur task is to detect where the diff lies.
--> task might be hard in part cuz the change is in an unexpected location. -->Ex, in 1 situation the engine of an airplane has been deleted—an unexpected change considering u have little prior experience with engines of planes disappearing! Yet once u detect the change in the engine, u can’t help but see it.
- Based on many past experiences we have expectations about what will occur in a given environment, & these expectations guide our attention
42
The value of a cognitive model lies in how well it explains the ________________.
available data
- The value of a cognitive model can be judged by how well it makes it explains data, but also how well it makes new, testable predictions.
43
The most intuitive evidence for Multi-Store Model comes from the ___________________
free recall paradigm
- participants r required to study of list of words presented 1 @ a time.-->They then need to recall as many words as they can in any order.
--> turns out that the words near the middle of the series r the least likely to be recalled while the 1st few words & the last few words r most likely to be recalled.
44
Memory performance is usually best for items that...
...were presented earlier or later in the list, and worst for items presented in the middle of the list.
- This pattern of performance is summarized by a graph like this, referred to as a serial–position curve.
- the multi-store model can explain this data.
45
the primacy effect
- memory performance is good for items encoded early in the list
why does this happen?
- According to the multi-store model, the items at the beginning of the list will be the first to enter short-term memory, & so will have the most opportunity to be rehearsed.
--> Ex, if ur task were to rmr items in the list “butter, tiger, hat, jump”, u might rehearse these items by repeating them to yourself in sequential order. --> cuz the 1st items in the memory list will be the most rehearsed, they have the best chance of being transferred into LONG-TERM
memory for more permanent storage.
- items in the middle of the list have less opportunity for rehearsal, & thus less
opportunity of being transferred into long-term memory. --> Consequently, the middle items r not rmred as well.
46
The Recency Effect
- The last items presumably have the least opportunity for
rehearsal, yet they are still more likely to be rmred than the middle items --> this is referred to as the recency effect.
why does this happen?
- According to the multi-store model, encoded info is
1st held in the short-term memory buffer that has the capacity to hold around 7 items
--> cuz you can only hold about 7 items in short-term memory, the newest items replace the oldest items as u go through the list --> the replaced oldest items r added to long-term memory
- when u reach the end of the list, the last 7 or so items r not replaced by new items & so remain in short-term memory --> basically memory performance is good for the last few items cuz they’re still active in short-term memory just before the memory test.
47
the primacy & recency effects r thought to diffly depend on the short-term & long-term memory stores. Explain
- the primacy effect is thought to reflect retrieval of words from long-term memory (after all, they were seen @ the very start of the list) whereas the recency effect is thought to reflect the immediate ‘readout’ of wtv is currently being rehearsed in short-term memory.
48
let’s hone in on another key assumption of the multi-store model. If short-term & long-term memory really do constitute 2 separate systems, then...
... we should be able to find variables that affect 1 without affecting the other, & vice versa.
49
Improving Primacy Effect
- We can manipulate how much time separates the presentation of each item in a
list. --> Having more time to encode each item would allow for more rehearsal & facilitate
transfer into long-term memory --> that's exactly what we find!
- Thus, the multi-store model predicts that the primacy effect should be more pronounced when there is
longer time between each item presentation.
- more rehearsal alone should not alter the recency effect cuz the words near the end of the series would still presumably be maintained in short-term memory to the same extent as when there is less rehearsal --> which is what happens --> primacy effect is usually observed (& the words from the middle of the series r also better recalled on average), but the recency effect remains unaltered
50
Diminishing Recency effect
- If the recency effect is indeed driven by items remaining active in short-term memory, then this result
should be influenced by a manipulation that removes or disrupts info in the short-term memory store before the final memory test.
--> We can test this prediction by having participants perform a distractor or filler task immediately after the
encoding phase, like counting down from 100 in increments of 3. ---> This distracting task requires short-term memory resources, & thus should replace any lingering items held in the short-term memory buffer.
51
Diminishing Recency effect: experiment setup
- an experiment involving 3 groups of participants:
1) The 1st group is asked to memorize the list of items, then performs a distractor task for 30 seconds before recalling the list.
2) The 2nd group is asked to memorize the list of items, then simply wait for 30 seconds before recalling the list.
3) The 3rd group is asked to memorize the list of items, then immediately recalls the list.
52
Diminishing Recency effect: experiment results
- As predicted by the multi-store model, the recency effect disappears when 1st
group is required to complete a 30 second distractor task that eliminates items from their short-term memory.
- the 2nd group who were left to sit in silence for 30 second interval, the recency effect is left completely intact --> that is, they performed similarly to the participants who immediately recalled the items from memory (group 3)
--> Why? Well, at no point were the items wiped from the silent interval group’s short-term memory-->they
were allowed to keep repeating the items as they had done up to that point & thus their performance
was as if they had experienced no gap at all.
--> Unlike the primacy effect, which can be strengthened or
weakened depending on manipulations to the memory task, the recency effect can only be diminished or remain as is.
53
Patient H.M & multi-store model: conclusion
- a lot of his behaviours begin to make more sense through the lens of the multi-store model.
--> His long-term memory system may have been impaired cuz he could not rmr things over the long-term, but his short-term memory system remained intact, as evidenced by his relatively normal behaviour across short time intervals.
--> However, memory researchers hold diff views about whether there truly is a dissociation between short-term & long-term memory, mainly cuz the boundary
between short-term & long-term memory can get fuzzy. --> For now, it should suffice to say that the multi-store model has been foundational & is still actively discussed today. --> the fact that we already knew terms like short-term & long-term memory is testament to this
54
Where r long-term memories stored in brain?
- We have learned a great deal about long-term memory from patients with acquired amnesia like patient H.M. --> can't rmr events from shortly before surgery (moderate retrograde amnesia), unable to form new long-lasting memories (anterograde amnesia) --> however memories from past were fully intact.
-->could follow short convos so long as they stayed within the range of short-term memory (~20 seconds) & could learn new procedural skills (e.g., a new set of motor actions, as one might need to play a new sport).
-Note that H.M.’s deficits were caused by a surgery that removed much of his hippocampus in the medial --> 1 might therefore think that long-term memories “reside” somewhere in the hippocampus.
- but this is oversimplified
- long-term memories seem to be represented in a distributed fashion throughout the cortex (outermost layer of the brain) & the gateway to the cortex is through the hippocampus which serves a critical role in the formation of these long-lasting representations.
--> our memory of a past event is reconstructed by co-activating some of those distributed parts
--> explains why H.M. rmr so much from long before his operation; these memories were already relatively well-consolidated in the cortex
--> also explains why he was unable to form new long-term memories; the hippocampus is critical for creating these long-lasting memories, & his was severely damaged.
55
Researchers also sometimes subdivide long-term memory itself. What 2 main categories is long-term memory divided into?
- Declarative memory & nondeclarative memory
56
long-term memory: Declarative memory
- aka explicit memory
Memories for factual info (semantic memory) or memories that r tied to a particular place & time (episodic memory).
- describes the more intuitive form of memory where the info is explicit & accessible to consciousness.
-->Ex, I ask what u rmr about eating for breakfast yesterday, & u tell me “avocado on toast”.
- Declarative memory is divided into semantic memory & episodic memory
57
Declarative memory: semantic memory
- general info that u can recall without necessarily rmring a specific time or place u learned it (e.g., that roses r red)
58
Declarative memory: Episodic memory
- info that u can recall about a discrete past event that is tied to a particular place & time (e.g., how red the roses were that sunny day at the dog park).
59
long-term memory: nondeclarative memory
- aka implicit memory
A combo of implicit & procedural memories. --> These memories r more automatic & unconscious.
- describes actions, procedures, skills, & conditioned responses--> info that is learned but often hard to articulate in words.
--> Ex, u know how to ride a bike, but good luck describing to someone else all the individual motor actions required to do so
- Long-term memory thus has a history of being decompartmentalized based on the types of info being retrieved.
60
Declarative memory vs nondeclarative memory
- Declarative Memory: explicitly rmring specific facts or episode.
- Nondeclarative Memory: implicit & procedural memories that r not overtly recalled but still drive functioning.
61
there r several issues with dividing memory into separate stores.
- tho the primacy & recency effects provide some support for the multi-store memory model, this doesn't necessarily mean that human memory is really organized this way
--> It's unclear just how info would be transferred between stores given the infinite # of ways that info can be rehearsed, & whether info must necessarily pass through short-term memory to get to long-term memory.
- also unclear to what extent short-term memory is capacity limited (1 of the features that distinguishes it from long-term storage) cuz short-term memory can the 7±2 capacity through chunking--> & a ‘chunk’ can only be identified as a result of its memorial consequences, thereby creating a circular explanation
- suffice to say that not all researchers agree about the multi-store model or whether memory should be conceptualized as being ‘carved up’ into all these discrete subsystems
62
General idea behind the levels of processing principle
- main idea is that encoding an item at a shallow level requires little effort, & is often directed @ the superficial physical characteristics of a stimulus. --> As a result, subsequent memory performance is poor.
- Encoding items at the
acoustic, or moderate level, fittingly leads to moderate memory performance.
- encoding an item at a deeper level requires more effort and is directed at the semantics or meaning of a stimulus. --> As a result, subsequent memory performance is better.
63
testing the levels of processing model
- In an influential experiment, Craik & Lockhart (1972) tested the levels of processing model by manipulating how participants encoded a list of words.
---> those in the shallow encoding group made physical judgments about whether the word was
printed in upper- or lowercase.
--> those in the moderate encoding group made acoustic judgments about whether the word rhymed with another word
--> those in the deep encoding group made semantic judgments about whether the word was related with another word
- After the encoding session, participants were given a surprise recall test. --> The results aligned with the
levels of processing model: memory performance was poor for the shallow level items, improved for the
moderate level items, & was best for the deepest level items. (i.e., meaning > phonological > physical).
64
levels of processing
A proposition by Craik & Lockhart that memory is not subdivided into separate stores. --> Instead, memory is conceived of as a continuum; items encoded @ a deeper level (more attention, more elaboration) result in a longer lasting memory trace than items encoded @ a more shallow level (less attention, less elaboration).
- emphasizes the importance of the encoding processes that underlie the formation of a memory trace to begin with.
65
The Levels of Processing Principle
The more we try to organize & understand the material, the better we rmr it
66
The Levels of Processing Principle: Criticisms
- 1 major criticism is that it pertains mostly to simple word stimuli.--> it's easy to say when a task requires u to focus on the meaning of a word, but it is hard to scale up the same logic to more complex learning tasks
--> Ex, think of this course's content--> what does it mean to study “deeply” in terms of meaning? -->Can 2 diff study strategies both target meaning, but 1 still lead to better learning than the other?
--> At this scale, it’s clear that the notion of “deep encoding” becomes ambiguous
- a 2nd criticism is that the logic becomes circular.--> wtv study methods lead to better memory will be considered deep & whatever methods lead to worse memory will be considered shallow.--> this
does little to pinpoint why 1 approach leads to better memory performance than the other.
- These criticisms aside, ideas related to levels of processing r related to some evidence-based study strategies--> Ex, periodically testing urself on the content, speaking aloud as if teaching urself a concept, & actively making connections with existing knowledge --> All these strategies can be thought of as “deep” in that they require effort & usually lead to better learning than more passive strategies.
67
Summary: Shallow level vs moderate level vs deeper level
Shallow Level
- Encoding requires little effort
- encodes physical characteristics
- poor memory performance
- shallow= physical
Moderate Level
- encoding requires some effort
- encodes acoustic characteristics
- moderate memory performance
- moderate= acoustic
Deeper Level
- encoding requires significant effort
- encodes semantic characteristics
- better memory performance
- deep= semantic
68
Mnemonic strategy
A device such as a pattern of letters, ideas, or associations that assists in rmring something.
--> One such mnemonic strategy is to think about the to-be-rmred info as it pertains to u
--> Generally, info encoded with ‘me-in-mind’ is better rmred than info encoded with something or someone else in mind=the self-referent effect
69
Self-Referent effect
Info encoded with ‘me-in-mind’ is better rmred than info encoded with something or someone else in mind.
--> seems to be quite a potent way to encode info, cuz words processed according to ‘self’ r better rmred than words processed by either meaning, phonology, or physical characteristics
- essentially, self > meaning > phonological > physical
70
Encoding Specificity
- in real world, memory is not restricted to learning lists of items or isolated bits of info.--> Instead, memory encompasses a rich & detailed experience of the world around u.
--> means that some environmental cues r incorporated into memory along with the studied
items= the principle of encoding specificity.
--> principle suggests that even when u encode a word in a memory experiment, u r encoding all of the specific aspects of that experience including properties of the room, the chair u r sitting on,
the font that the words r written with. -->cuz some
environmental cues r stored in memory, they can act as retrieval cues to help access the og studied info in memory.
- thus, it can be beneficial to tailor ur study environment (both physically & mentally) to be as similar as possible to the future testing environment. --> when studying, set up ur desk like an exam table & get mentally prepared by putting away ur notes, turning off ur phone, & timing urself.
- Better yet, try to study the info in multiple diff contexts. -->Theoretically, the more contexts the info is linked to, the more ‘paths’ there r in memory to retrieve the info.
71
Encoding specificity textbook def
- When we encode a particular item into memory, the item is not processed in isolation but together with the surrounding context.
--> The effectiveness of the context as a retrieval cue is largely dependent on the specificity of its relations to the item--> to what extent the retrieval cue selectively brings to mind the target info n while also excluding nontarget info.
72
What is one implication of the encoding specificity principle?
- memory performance ought to be better when the
conditions during retrieval closely match those of initial encoding
--> in a demo, Godden & Baddeley (1975) asked SCUBA divers to encode a list of words while on land (beside a pool) or while underwater (in the pool). --> A follow-up recall test was done either in the same or diff encoding context.
--> participants were better able to rmr items from the
list when they were tested in same context that was present during encoding.
73
Encoding Specificity principle: Context can also be conceptualized as a person’s _______________________. Elaborate
internal or mental state.
--> Ex, Eich & colleagues (1975) conducted an experiment where they asked participants to encode a list of words after smoking cannabis or after smoking a placebo, then to complete a recall test after a delay, before which
they smoked either cannabis or a placebo
--> Interestingly, participants who had previously studied while under the influence of cannabis rmred more on the recall test if they were also under the influence of cannabis while writing it,
when compared to if they received the placebo before writing it.
--> However, keep in mind that the best performance was observed in the placebo–placebo condition!
74
particular manner of encoding a stimulus also acts as a strong form of context. Explain
- rmr Hermann Ebbinghaus who used himself as a subject by memorizing strings of nonsense words --> he chose these nonsense words to minimize the influence of prior experience on his test results.
--> What he found was that each word in the list served as a cue to trigger the memory of the word that followed; like pearls linked in a necklace, each word connects to the word before & after it.
--> u can think of encoding context as the other words that were encoded shortly before or after the target word. --> Ex, if u encountered the series of nonsense words “tol, dal, puh” somewhere in a list during encoding, u would probably have a much easier time recalling “dal” if context is reinstated by being given the retrieval cue of “tol”.
75
Transfer-Appropriate Processing
Memory is aided when similar mental processes r engaged @ encoding & retrieval.
--> This is primarily due to encoding specificity
- 1st showed this using a levels of processing paradigm
--> On each study trial, a sentence with a blank space appeared, then the participant indicated whether a target word could replace the blank to create a meaningful statement.--> There were 2 trial types: semantic (“The ___ has a silver engine.”) & rhyme (“___ rhymes with legal.”)
--> Afterward there was a surprise memory test.
- 1 group did a standard recognition task, trying to discriminate ‘old’ words from ‘new’ words. --> This group’s performance aligned with the levels of processing framework--> words processed in semantic sentences were better rmred than words processed in rhyme sentences.
- other group performed a special rhyme recognition task where they determined whether each word rhymed with any previously presented word.
--> In contrast to the levels of processing framework, this group showed better recognition for target words previously presented in rhyme sentences than in semantic sentences. --> Thus, invoking similar processes during both encoding & retrieval (i.e., phonological processes associated with rhyming) was sufficient to overshadow the typical benefits of deep semantic encoding.
76
Forgetting: Ebbinghaus
- Ebbinghaus was also very interested in how long memories could be maintained.
--> Unsurprisingly, he found that his ability to recall words was best immediately following learning, & that he was able to rmr fewer & fewer words with longer retention intervals until performance reached an asymptote close to 0.
- critically, he found that the rate of forgetting started high & gradually decreased. --> Info thus seemed to decay from memory very rapidly shortly after the initial learning took place, but then the rate of forgetting per unit time decreased from that point forward.
--> We therefore refer to this as a negatively accelerating forgetting curve.
- These experiments by Ebbinghaus were the 1st to illustrate the rate of forgetting.
77
Forgetting curve
memory savings plotted over time
- a negatively accelerating function
78
Forgetting: Decay theory
- This is 1 way that memory researchers think about forgetting
The idea that forgetting occurs cuz memories naturally fade over time.
--> tho decay does a decent job of describing forgetting in short-term memory, it does not adequately explain forgetting in long-term memory cuz the passage of time is not always a good predictor of recall.
--> also, an event that seems to have been completely forgotten can often later be rmred given the right retrieval cues.
79
Forgetting: Interference
- a diff way to think about memory is through interference= an inability to access info from memory due to other SIMILAR INFO "competing" for retrieval.
--> Another way to think about this is in terms of competition between the similar sources of info.-->@ some later time when u go to retrieve relevant info on one topic, info from the other topic might compete for retrieval, thus preventing u from recalling the target memory.
--> Indeed, many studies suggest that similar forms of interference can explain some aspects of forgetting,
especially for LONG-TERM memory.
Textbook
- many theorists argue that stored memories r basically permanent & that forgetting is merely the consequence of not having the right retrieval cues to access them-->Ex, a cue that was once useful for retrieving a memory may have since become associated with diff info, & consequently may no longer be able to retrieve the og memory. --> Such explanations of forgetting rely on the idea of interference
80
For adaptive purposes, we forget some of what we learned in the past, but memories r not necessarily __________
deleted
- we may have just lost the ability to retrieve them.
81
How can u overcome interference?
--> Ex, u sit down to study a topic like animal physiology. A few hours later, u sit down again to study a related topic, cell physiology.
--> A few days later, u will write a test specifically on the prior animal physiology content. --> Beyond the mere passage of time, does the latter studied content overshadow the recall of the previously studied content?
--> if u r simply cramming the study of the 1st topic, followed by cramming of 2nd topic, u r setting the stage for interference.
--> u can overcome possible interference with memory recall by implementing a study plan to initially build strong memories with many retrieval routes.--> Actively engaging with study material & building connections with ur existing knowledge & experiences will strengthen later recall & minimize the impact of possible interference.
82
Proactive interference
Info learned prior to a specific memory interferes with its retrieval.
--> Ex, u learned French in high school & now @ uni u want to learn Spanish. --> tho ur previous knowledge is helpful, sometimes when trying to rmr a specific Spanish term u rmr the corresponding French term instead. --> ur old memory for high school French is interfering with the retrieval of ur newer memory for Spanish words.
- We can measure proactive interference in the lab using the following simple paradigm. --> experimental group studies word list A & then immediately studies word list B. -->control group only studies word list B.
--> After a brief interval, both groups recall words from list B. --> Proactive interference has occurred if the experimental group rmrs fewer words than the control group.
83
Retroactive interference
Info learned after a specific memory interferes with its retrieval.
--> Ex, experiences of what u ate several times since then interferes with ur retrieval of what u at last Wednesday
- In lab, we can measure retroactive interference using a similar paradigm to one used for proactive interference --> A control group only studies word list A & then plays Tetris (or does some other filler task).
--> The experimental group studies word list A & then immediately studies word list B. --> After brief interval, both groups recall words from list A.--> Retroactive interference has occurred if the experimental group rmrs less words than the control group.
84
Interference vs. Decay Models: Rugby players
- experiment conducted by Baddeley & Hitch (1977) that tested the memory of rugby
players.
--> According to the decay model, time alone ought to predict how well the players remember details about
prior games.
--> In contrast, according to the interference model, it’s the # of similar experiences,
the # of intervening games, that should acc predict how well players rmr details about prior games.
--> The researchers could disentangle the contributions of time & the # of intervening games played cuz some players had to sit out some games due to injury.
- Their key finding was that the strongest predictor of forgetting was the # of intervening games played & not time per se, consistent with the interference model & not the decay model.
85
Interference vs. Decay Models: conclusion
- We usually think of memory in terms of decay over time
--> But, as implied by interference models, things
become more intriguing if we acknowledge that forgetting may be more about an inability to retrieve something stored in memory rather than passive decay.
--> @ 1 extreme, researchers have conjectured that even the faintest of memories might still be stored in the brain in some way--> they r just rendered inaccessible by heaps of other interfering info.--> Unless very specific retrieval cues r presented to aid rmring, such stored memories may be functionally lost.
86
Repressed Memory
A painful memory that has been forgotten as a defense mechanism.
- According to Freud, horrific memories (like childhood abuse) can be repressed as a defense mechanism
- there is evidence to suggest that people can acc forget child sexual abuse.
--> Ex, Williams (1994) interviewed 129 patients who were medically treated for sexual abuse that had occurred about 20 years before. --> 38% of those interviewed failed to recall the abuse. --> Of those that had forgotten, many were under 6 when the abuse was originally documented.
--> But is this really evidence for memory repression?
--> It's possible that forgetting a horrific event is simply an ex of ordinary forgetting, & that instead we seem to have a general bias for rmring positive or pleasant events rather than unpleasant events
--> Ex, painful experiences like giving birth to a child are not well rmred.
- In short, the ans to this question remains controversial.
87
Why do we forget?
- As a final point, forgetting may not always be an enemy. --> It might be counterproductive to rmr everything about the past, especially when u know that some of that info does not
apply to the future.
--> Some info is worth rmring for immediate needs (ur hotel room # & where u parked ur car), but is not something u need to rmr long into the future.
- what would happen if u never forgot anything u learned; part of ur problem might actually be an excess of info stored in long-term memory, some of which may be useless or outdated. --> Ex, forgetting reduces the interference from such out-of-date or irrelevant memories, & may in turn allow us to rmr relevant or up-to-date info more effectively
- thus, there is an argument to be made about the process of forgetting being adaptive in some circumstances
88
False Memories: Elizabeth Loftus
- in experiment by Loftus, participants were sent a booklet containing several of their own childhood (ages 4-6) experiences that were gathered from interviews with their older relatives.
--> Of the 4 experiences presented to the participants, 3 were real, & 1 was fake, made up by the experimenter.
- The fake experience described a detailed memory of getting lost in mall, crying, receiving comfort from an elderly woman & a happy reunion with the family. --> Participants were asked to recall & describe these 4 experiences in the booklet, & then again later over 2 separate interviews spaced approximately 1-2 weeks apart.
- In the booklet, 29% of participants recalled the fake experience as being real, & only 1 of these participants corrected their mistake during the following interviews.--> thus, 25% of people believed a fake experience to be a genuine memory from their
past.
--> These false memories demo that memory is a highly reconstructive process.
- & cuz people can be strategically manipulated into believing an event occurred when it did not, we now know that eyewitness testimonies in the court of law can be unreliable.
89
Loftus & Washington Sniper case
- 2002, police were investigating the case of the Washington Sniper, a serial killer targeting random victims in Washington, DC.
- many eyewitnesses said that a white van was spotted near several of the murders, but no evidence of the white van was ever found, & the murderer’s vehicle turned out to be a dark green car.
--> How did this memory of a white van come to be?
- work from Elizabeth Loftus suggests that our memories r far from perfect & that we r susceptible to false memories. --> Loftus noted that an early eyewitness had participated in a media interview mentioning a white van. --> Evidently, this idea stuck, & subsequent eyewitnesses also (inadvertently) propagated the same false memory.
90
Misinformation effect
The creation of false memories by incorporating new erroneous info with an old memory.
--> demoed this by asking participants, all of whom had visited Disneyland as a child, to evaluate a new (fake) advertisement for Disneyland featuring Bugs Bunny.
--> tho Bugs Bunny is a Warner Brothers cartoon character & would never be associated with Disneyland, 16% rmred shaking hands with Bugs Bunny when they visited Disneyland as a child.
91
The language used during legal questioning can also intro new & inaccurate info. Elaborate
- Ex, Loftus & Palmer (1973) had participants watch a film of a car accident & then asked them: “About how fast were the cars going when they _____ into each other?”
--> Participants’ estimations of speed varied according to the verb of the leading question; the word “smashed” resulted in the greatest speed estimate, “collided” resulted in a lower speed estimate, & “bumped” resulted in the lowest speed estimate.
--> Further, the word “smashed” led more people to falsely rmr broken glass even though there was no broken glass in the acc film.
92
False memories: Bizzarre experiences experiment set up
- u may argue that from the Loftus study, the experience of getting lost in a mall may be an induced false memory, but it is a plausible false memory. --> Getting lost is a common enough experience
& perhaps this might influence the participant into believing it to be true.
- What if we considered more
bizarre experiences?
- study by Seamon & colleagues (2006) demoed that repeatedly imagining an event can lead to a false memory, even for very bizarre situations.--> Participants
visited various locations on a university campus & @ each stop, a familiar or bizarre action was either performed or merely imagined.
--> Ex, when a participant stopped at a vending machine, he was either asked to perform or imagine a familiar action, like looking for change. --> In another condition, a participant was asked to either perform or imagine a bizarre action like proposing marriage.
93
False memories: Bizzarre experiences experiment Results
- On a follow-up test 2 weeks later, having simply imagined that bizarre action just once, was enough for those participants to be more likely to claim that they had acc performed & not imagined the bizarre action.
--> Under these circumstances, it would appear that memory can be misled into confusing imagined & performed events no matter how bizarre they may be.
94
Reality Monitoring
Our ability to discriminate real memories from those that r thought or imagined.
--> Ex, the bizarre actions experiment
95
False memories provide more support for the idea that memory is a________________ ___________. Explain
reconstructive process.
- Rather than providing an accurate snapshot of reality, recalling a memory is open to interpretation & suggestion.
- Think about how the details of ur favourite stories u tell change slightly as u tell them over the years & from person to person.
96
Fluency
The ease with which an experience is processed, some easier (more fluent) than others
- Familiar experiences r
generally processed more fluently than r novel experiences.
--> The sense of familiarity that u felt when looking @ the stranger on a bus is a kind of processing fluency.
97
Attribution
Judgement tying together causes with effects
- We can make an attribution to the processing fluency of the feeling of familiarity to the stranger on the bus, which can potentially be tied to many diff causes.
98
Fluency can be a double-edged sword. Explain
- it can act feel like a superhero power, a sort of "spidey sense“ that makes a task feel like it was processed easier
--> Most of the time we do this seamlessly, often
attributing our feeling of fluency to a familiar experience or an easily recalled memory.
- That’s where we have to be careful about our own metacognitive awareness of assigning the source of
the feeling of fluency, which may not be accurate.
--> ex, we could attribute the feeling of fluency to mean that we must know this person from somewhere, but we could also make other attributions. --> Perhaps we could instead think that this person stands out cuz they look unconventional, or just sitting @ the front of the bus
--> There r many stories that people can create in order to interpret their internal experience of fluency
99
Becoming Famous Overnight: Set up
- An experiment entitled “becoming famous overnight” provides an interesting look @ processing fluency in action.
--> Phase 1: participants were asked to read aloud a list of people's names in a
pronunciation task
- Participants proceeded to Phase 2 immediately or after a 24 hour delay.
--> Phase 2: participants viewed a 2nd list which contained the names of celebrities & generic names
that had been made up for the study. --> They were asked to give a fame rating for each person on this 2nd list.
--> some of the made-up names were new & others were from drawn from the
same list that participants had seen in the 1st phase
100
Becoming Famous Overnight: Results
- Not surprisingly, participants generally agreed upon the fame rating of the names of the acc celebrities.
- Interestingly, the time delay between phase 1 & 2 was important.
--> Participants who proceeded directly from the pronunciation task in phase 1 to the fame rating task in phase 2, appropriately rated the names of celebrities as being famous & the generic names as not.
--> but for or participants who waited overnight before phase 2, participants rated the generic names encountered the day before as being more famous than the new generic names!
101
Becoming Famous Overnight: Results explanation
- consider things from the participants’ perspective.--> Imagine u come across a generic name & u experience processing fluency.
--> This fluency can be explained by making the attribution that the name in question must be at least somewhat famous --> this is a
misattribution generated to explain processing fluency
- cuz of the long delay, it may be more difficult to properly attribute the processing fluency to the fact that u have previously been exposed
to the name the day before
- For the group that immediately proceeded to phase 2, a participant can
more easily make the correct attribution that fluency with the name in question is attributed to the fact that they had just seen this name in the pronunciation task.
102
In sum, you can think of memory as a...
... large collection of raw materials or building blocks that were assembled through perception & experience
- But each event is represented by multiple building blocks to represent its many parts, including things like the stimuli, one’s internal state, or the physical
environment.
--> And although rmring is often an attempt to reconstruct the building blocks of 1 specific event, this process is imperfect. --> Perhaps cuz some of the blocks cannot be accessed, as in forgetting, or blocks from other events r falsely incorporated, as in false memories
103
False fame effect
We attribute fame to a name since we cannot rmr where we have seen it before (source monitory error).
--> Since we recognize the name, it must be famous!
104
The idea of a true or objective memory is _____________. Elaborate
impossible
- rmring is not a reliable rehashing of the past, but an inference--> it's simply the brain’s best guess at what
happened based on everything else u have stored in memory
--> & once u appreciate this complex process, it should baffle u just how often these “best guesses” do their job well enough to help us
navigate through this world!
105
Source Monitoring Error
When we cannot recall where we learned a piece of info.
--> The process by which we make attributions about the origins of memories= source monitoring
- 1 way in which we commit a source monitoring error is by misattributing fluency (i.e., the subjective ease of processing) when we experience something familiar.
106
Attention works to selectively process info that is... Elaborate
...salient or most relevant to our goals.
--> Without attention we couldn’t listen to a friend in a noisy bar, find a colleague in a crowd, or notice important events in the environment like an ambulance siren.
- Though attention makes us more efficient, attention is also fallible & sometimes our expectations lead us to miss highly salient items.--> Working together, attention & memory shape our experience of the world around us.
Psych 1X03
flashcards
Decks in class (13)
# Cards
-
Research methods 1107
-
Textbook: Intro: foundations of psych11
-
Classical Conditioning82
-
Instumental conditioning69
-
Problem Solving & Intelligence65
-
Language82
-
Categories & Concepts57
-
Science of learning22
-
chapter 7 textbook: Attention0
-
Attention!73
-
Influence of others oney158
-
Forming Impressions (HEHE)69
-
Memoryyyyyyyyyyyyy106
