NEUR1020 - QUIZ 3

what is learning

learning is defined as the modification of behaviour through experience

what did Eric Kandel discover

• foundational insights into very simple forms of learning were achieved through Eric Kandel’s work with the sea slug, aplysia californica 

• Kandel’s work demonstrated that the strength of the gill withdrawal reflex can be modified with experience

how can we demonstrate Kandel’s work through habituation

• through habituation, we apply a jet of water to the aplysia’s siphon and the difference between the initial measurement and the post-reaction measurement gives us an index of the reflex strength

• we continue to disturb the slug every 90 seconds and find that the strength of the reflex gradually weakens due to repeated exposure

how can we demonstrate Kandel’s work through sensitisation 

• in sensitisation, instead of continuing to apply the water jet stimulus, we switch to a new, more intense, stimulus (ie. electric shock) 

• we find that the strength of the reflex increases

• we then reapply the water jet stimulus and find that it elicits a stronger reflex compared to the first

• this is strengthening of the reflex to the same stimulus 

what is habituation

shows how a relatively important stimulus comes to be ignored ( decreasing strength)

what is sensitisation

shows how an organism becomes more vigilant about a stimulus (increasing strength)

habituation and sensitisation are both highly adaptive ways of allocating resources or prioritising different behaviours 

what was pavlov’s study

• looked at how associations are formed through investigating dogs and salivations responses to food and anticipatory cues (ie. bell)

what is classical conditioning

classical conditioning is when a previously neutral stimulus elicits the same response as another stimulus, after repeatedly being paired together

any study of classical conditioning involves 4 key elements

what are the 4 elements of classical conditioning

1. unconditioned stimulus (ie. meat)

2. unconditioned response (ie. salivation) → denotes that the animal does not need to learn how to respond it is innate

3. conditioned stimulus (ie. metronome)

4. conditioned response (ie. salivation after repeated pairing) → denotes that the response is learned via association of repeated pairings with the UCS

what is high order conditioning

• a conditioned response can be generalised from one conditioned stimulus to another 

• we do this through pairing the CS and UCS to train a CR. Then later we pair the original CS with a new CS. The new CS will elicit a weakened CR by itself

what was Thorndike’s law of effect 

• one of the earliest demonstrations that animals could learn to associate behaviours with consequences 

• this was done by a puzzle box where a door could be operated by a cat

what is operant conditioning

• based upon the principle that behaviours that are followed by satisfying outcomes tend to be repeated over unsatisfying outcomes

• in operant conditioning, learning depends on:

1. the behaviours of the organism

2. the consequences of those behaviours to the organism

what was skinner’s box

• done to rigorously monitor behaviour and controlled administration of both pleasant and unpleasant consequences 

what are the main components of operant conditioning

1. reinforcement refers to the outcome that increases the strength/frequency/probability of a behaviour

2. punishment refers to the outcome that decreases the strength/frequency/probability of a behaviour 

what makes an outcome reinforcing or punishing depends on what has been added or removed from the environment 

what are the 4 classes of outcomes in operant conditioning

1. positive reinforcement → produced a pleasant consequence to behaviour by adding a stimulus

2. positive punishment → add something unpleasant to behaviour

3. negative reinforcement → remove something unpleasant when behaviour is performed

4. negative punishment → remove something pleasant to minimise behaviour

what are schedules of reinforcement

• with operant conditioning, we need to consider how consistently reinforcement is being delivered to the learner:

• continuous reinforcement → every instance of the behaviour is reinforced

• partial reinforcement → only some instances of the behaviour are reinforced (this tends to produce more robust learning)

time vs response based schedules of reinforcement

• reinforcement can be based on either time or responses:

1. interval schedules → time based reinforcement (ie. weekly paycheck)

2. ratio schedules → response based reinforcement → amount of reinforcement scales directly with the rate of responding (ie. coffee loyalty card)

fixed vs variable interval/ratio schedules

• both interval and ratio schedules can be fixed or variable:

1. fixed → reinforcement is provided regularly (predictable)

2. variable → reinforcement is irregular (ie. time wise or number of responses)

variable schedules produce steady responding. Fixed schedules produce more erratic ‘stop-’start patterns of responding when we plot the cumulative number of responses as a function of time

what is the classical conditioning principle of acquisition

the first principle is acquisition. Whereby a conditioned response does not appear instantly. Rather it is gradually acquired by repeatedly pairing the UCS and CS.

Early learning is characterised by more rapid trial-by-trial changes in the CR. Later on in learning, there is progressively less change in the CR, as it approaches the asymptote

acquisition is time sensitive. Both the rate and the asymptotic strength of the conditioned response depend on the relative timing of the CS and the UCS

learning is most efficient when the CS is presented shortly before the UCS. Backward conditioning is where you present the UCS before the CS → however this is often ineffective. This may be due to potential evolutionary significance → a cause must occur before an effect

what is the classical conditioning principle of extinction

the second principle is extinction. Whereby a conditioned response will reduce in strength and eventually disappear if the CS is repeatedly presented without the UCS:

extinction is a common way of eliminating a CR. Another way is to try and pair the CS with a new CR.

the previously trained CS-CR association remains intact → potentially being suppressed by other more active associations.

presentation of the original learning context can trigger spontaneous reappearance of the CR. The CR can also be brought back if a similar original stimulus is introduced

why is extinction not the same as forgetting 

important to note that extinction is not the same as the dog forgetting a CS-CR pairing. This is because of spontaneous recovery. If extinction produces forgetting → it should not be possible for a CR to reappear unless the CS-CR association has been retrained. However, we often see a reemergence of the CR.

how similar must the stimulus be to bring back original CR (principle of generalisation and discrimination)

Stimuli that are similar to the CS will tend to elicit the same CR as the CS itself. The blue curve shows that when this happens, the CR has generalised to the novel stimulus. Whatever the animal learns about the original stimulus tends to be carried on through to the new stimulus

the red curve shows that if the stimulus is dissimilar from the CS, it will not elicit the CR. The organism is able to discriminate the CS from the new stimulus

what are some applications of classical conditioning

• everyday conditioning occurs within advertisements (ie. alcohol ads and refreshment, enjoyment, and attractive people) → advertisers pair their products with an UCS to elicit the same feelings/thoughts as the UCS.

• another application is fears, phobias, and little Albert → followed the ideas of behaviourism. Controlling the environment children grow up in so that they learn to be firm.

• some other applications of classical conditioning is behavioural treatments for phobias. We wish to extinguish an association between the CS and CR, for example desensitisation technique

• classical conditioning is also linked to disgust and taste aversion. Disgust towards certain stimuli can be viewed as a form of classical conditioning → this response can generalise very readily even to things the person claims they like. It is a strong ‘one trial’ learning after food poisoning experiences.

what was the little Albert experiment 

• A notorious study conducted by Watson to show that fears were learned behaviours; not manifestations of unconscious conflict → little Albert shows no fear of rat → the initial part of the experiment shows that the rat is a neutral stimulus. However, the UCS was a gong, and eventually the rat becomes associated with the fear/startle response. However, the fear response generalised to other similar stimuli (ie. dogs).

• however, this experiment was controversial because there was no opportunity to de-condition little Albert.

why is reinforcement methods more effective than punishment (in operant conditioning)

• reinforcement tends to train a target behaviour more effectively than punishment → reinforcement is more focused and informative whereas punishment can lead to unexpected outcomes

• another issue with punishment is that if a punishment outcome can be anticipated by the learner, they may learn ways to avoid it.

what are the principles of operant conditioning

• also has the principle of extinction → eliminating a previously reinforced behaviour by no longer delivering reinforcement

• and it also has the principle of spontaneous recovery → reemergence of a previously reinforced behaviour despite suspension of reinforcement

• also has the principles of generalisation/discrimination → whether a reinforced behaviour is emitted in response to similar but not identical stimuli (honey bee experiment)

• the principle of acquisition is the same → incrementally associating a behaviour with an outcome (timing rules for classical conditioning still apply for operant conditioning)

• however, with operant conditioning, we also need to consider how consistently reinforcement is being delivered to the learner: (either consistent or partial reinforcement)

what is humphrey’s paradox

Humphrey’s paradox comes up as a comparison between partial and continuous reinforcement.

Schedule A (consistent reinforcement) produces faster learning (ie. faster acquisition). However, schedule B (partial reinforcement) produces more robust learning (ie. harder to extinguish)

what are the applications of operant conditioning

• one main application is shaping and animal training (Ie. getting a dog to sit) → to train a target behaviour, begin by reinforcing behaviours that vaguely approximate the target behaviour. Then, restrict reinforcement to behaviours that are increasingly similar to target behaviour.

• another application is phobias → fears that are acquired through classical conditioning may not passively resist extinction or forgetting. Anxiety caused by the conditioned stimulus can be relieved by actively avoiding the stimulus. Reduction in anxiety provoked by avoidance behaviour negatively reinforces the fear.

• operant conditioning can maintain superstitious behaviours (or rituals) via positive and negative reinforcement

what is memory

• memory is the retention of information over time 

what are the three different memory stores and their overall purpose

• the processes of encoding, storage, and retrieval all form part of the interactions between the three different memory stores:

1. sensory memory

2. short-term memory

3. long-term memory

what is the multi-store model

• we do not possess a unitary memory system. Rather memory comprises a number of interacting memory stores that each have their own characteristics

• as information progresses through these stores, the information that is retained becomes more robust but can be lost to forgetting along the way

what is sensory memory

• sensory memory is where there is a memory store for each of our sense modalities

• perceptual information is stored in separate streams but can be consolidated into a more durable short-term memory

• sensory information must be attended to and transferred to short term memory to avoid being lost to forgetting (bombarded with more sensory information than we can remember)

what is the time frame and capacity for sensory memory

• the time frame for this information transfer is quite brief:

1. visual store (iconic memory) → duration <1 second and capacity 12+ items

2. auditory store (echoic memory) → duration 5-10 seconds, capacity 5 items.

what is short term memory

• short term memory is not modality specific. Its contents are often aligned with the current contents of awareness

• information in short term memory is also information that is available for report 

• information held in short term memory will be lost to forgetting. However, it is possible to refresh short term memory through a process called rehearsing

What did Miller find about short term memory in relation to capacity

• Miller found that people are adept at holding around 7 distinct items in awareness at once

• measured through a span task → where you study a list of letter presented at a rate of around 1 per second and you have to recall as many digits as you can in the order they were presented and the list length when recall errors occur reflects storage capacity

what is rehearsal

• rehearsal involves mentally repeating information that is currently being held in memory, which is tantamount to representing the stimuli anew.

• rehearsal also facilitates the transfer of information to long term memory

what is the time frame of short term memory without rehearsal

• when experimental measures are taken to prevent rehearsal, the duration of short term memory is around 20-30seconds

what is long term memory

• any information we have retained for 30 seconds or longer (it is our sum of all knowledge)

what is retrieval

• retrieval is the process of locating information in long term memory and bringing it into short term memory so it can be reported

• retrieval of information is often supported by cues that are available in short term memory

what does the effectiveness of retrieval determine

• the effectiveness of retrieval process is the principal factors that limits how much information we can store in our long term memory 

what are the two capacity conclusions in regards to long term memory

1. capacity of long term memory is massive

2. once information is transferred into long term memory that information will persist indefinitely

how can we find the mechanisms responsible for forgetting

• if rehearsal is prevented we can isolate the mechanisms that are responsible for forgetting in short term memory

what are the two mechanisms responsible for forgetting

1. time based decay

2. interference (most supported mechanism)

what is time based decay

• information in memory is progressively degraded by the simple passage of time

• it follows that the extending of the retention interval will produce poorer results

what did peterson and peterson test

• they tested time based decay in a seminal study by making participants commit a letter triplet to memory

• however, some people argue that the key manipulation (counting backwards in three) does not just extend the retention interval but also create a distractor item restricting retrieval

what is interference

• the more distracting information there is, the harder it is to pinpoint and report target information

• build up of interference produces poorer memory performance

what is the free recall task

• in the free recall task stimuli are presented and reported back (ie. recalled) in any order

• the list items are monosyllable words often presented at a rate of one item per second

what is the serial position curve 

• from the free recall task we can measure recall accuracy as a function of list position/order → this is the serial position curve 

explain the shape of the serial position curve

• accuracy is nearly 100% for the last item and steadily drops off as you go further back into the list. As you approach the end of the list there is an improvement in performance (this is called the recency effect)

• in the middle of the list, accuracy appears to bottom out and stabilise

• for the start, performance actually becomes increasingly accurate → this improvement in memory performance is known as the primary effect

• the primary and recency effects give the serial position curve a highly distinctive U shape

what is the first theory that explains the primary and recency effects (and the limitation to this theory)

• in terms of the multi-store model, the very earliest items are encoded into the short term memory as they are presented

• the time between the presentation of successive list items can be used to rehearse that item information and this process is repeated until short term memory capacity is reached

• the first way to ensure that there is room in the short term memory to accommodate more recently presented items was thought to be to forget older items

• however, this approach does not explain the serial position curve shape or the primary effect

what is the second theory to the primary and recency effect

• if early list items are rehearsed, they can be transferred to the long term memory, freeing up space for new items

• early list items are able to be rehearsed more thoroughly → explaining the primary effect

• most recently presented stimuli are currently stored in the short term memory and the early list items are stored in the long term memory due to better rehearsal

• the middle list items were lost to forgetting as these items had the fewest opportunities to rehearse

what is a cue

• a cue is anything that serves as a signal for something else in relation to memory retrieval (ie. features of the environment that direct us to information in memory)

• cues become associated with target information and boost the likelihood of successful retrieval as remembering a cue can trigger memory for the target information

why are cues important

• having a cue associated with a memory target is like having an extra opportunity to retrieve the desired information

• having more reliable cues associated with the target increases the likelihood of retrieval further

what did gordon and batali find

• they illustrated how cues improve memory in their study on encoding context on cue context

• their graphs showed a clear benefit of matching encoding in retrieval contexts (land and water)

what is the encoding specificity principle

• retrieval is facilitated when cues that are present at encoding are also present at retrieval 

• note: incidental features of the environment can also become associated with target information

what did Craik and Lockhart find

• they introduced the depth of processing framework

• this suggested that stimuli that are processed in greater depth at encoding enjoy a memory benefit over stimuli that are processed more superficially

• this is because deeper encoding helps to link information up with other related concepts that can facilitate retrieval

give an example of cues to action

• overtime certain road features (ie. a T road or fork) will become strongly associated with certain behaviours such as turning on the indicator.

what causes interference errors

• when sometimes if an aspect of the environment changes, cues that used to signal appropriate responses can signal inappropriate responses resulting in interference errors

what are the two forms of interference errors

1. proactive interference → older information disrupts retrieval of more recent information

2. retroactive interference → more recent information disrupts retrieval of older information (ie. learners of a second language)

• it is not possible to predict whether interference will occur proactively or retroactively → it is dependent on context.

what are some theoretical implications of the multi-store model

sometimes we forget information if it is not transferred properly. However, when information is successively transferred to different memory stores → increasingly durable forms of memory.

The several points where information can be lost is: transfer from sensory memory to short term memory or transfer from short term memory to long term memory (if we don’t help to ensure consolidation or pay attention to current information)

what did sperling find in relation to sensory memory

• Sperling contrasted performance on full report and partial report versions of a letter memory task → full report performance was around 5 letters and partial report performance was virtually perfect

• this suggests that a large amount of information stored in visual sensory memory, but only for a short amount of time (otherwise full report would be as good as partial report)

what did the yearbook study find in relation to long term memory capacity

• the duration of long term memory was tested through the yearbook study → the interval between graduation and testing ranged from 2 weeks to 57 years → they found that matching performance and recognition of partial information is very good, however, free recall and cued recall fared worse, the only major drop off occurs after a retention interval of nearly 50 years.

• essentially it suggests that once information lands in long term memory stays intact for years

what are the two types of retrieval 

• there are two types of retrieval:

1. cue based retrieval → mnemonic strategies

2. beyond pure retrieval (memory as a reconstructive process) → schemas, misinformation, and false memory

what are some common mnemonic strategies for cue based retrieval

1. method of loci → associate memoranda with locations/landmarks along a familiar route

2. pegword technique → pairing vivid imagery with memoranda (ie. one is a bun, two is a shoe)

3. keyword method → pair a word that sounds similar to the to be remembered target.

what is the chunking strategy for retrieval

• we consistently remember phone numbers and credit card numbers → we do this by grouping the numbers into chunks (ie. 2 groups of 4 etc.) → this is a common technique used

• when we group we form an association between numbers. Elements within a group become associated with each other → they can form cohesive units and retrieval of one parts triggers retrieval of the whole unit

what did chase and ericsson find in relation to chunking

• chase and ericsson tested one participants for 260 hours over 2 years and found that with repeated practice improved is recall from 7 items to 82 items using complex chunking.

• this was possible because SF remembered sequences of numbers in terms of running times in different races.

what did Bartlett investigate with regard to cues and retrieval

Bartlett investigated cues and associations through people remembering and then drawing a picture of an owl. The owl image eventually turned into a black cat → this demonstrates that memory is not just about retrieval but also about interpretation

what are the three explanations to Bartlett’s study findings 

1. schemas → templates or scripts for familiar situations → if the details of something are forgotten, the default offered by the script can be inserted into one’s memory for what happened (shown in the allport and postman study)

2. subtle influences of misinformation → shown in the study of loftus and palmer indicated that wording can impact how people rate the how fast a car was going (ie. smashed, hit, collided etc.)

3. implanting false memories → this was shown on loftus and pickrell where relatives of participants provided experimenter with information on 3 events, and a fourth story was introduced by the experimenter → around 25% of participants reported remembering this fake story

what are the implications of cues and retrieval 

• this has implications for eyewitness memory as people may not accurately remember events or introduce fake events → thus multiple factors erode confidence in eyewitness memory (schema, and suggestibility)

• reconstructive nature of memory demands caution on the part of law enforcement agencies and jurors

what is a concept

• concepts are mental representations that organise our knowledge about specific things

• we understand the world through concepts

• concept knowledge supports recognition of objects and allow us to infer their use

what are concepts shaped by

• concepts are shaped not just by our experiences in the world but also through the cultural milu we find ourselves in

• it follows that the meaning we attach to things will differ across individuals and cultures even if we perceive them in the same way (ie. the meaning of constellations across cultures)

what is the classical view of concepts

• the classical view of concepts was proposed by Aristotle and says that concepts possess a definitional structure

• the features of this definition are:

1. something is either an example of the concept or is not

2. positive examples possess necessary features

3. possessing defining features is necessary

what are the two problems with this view

1. generating definitions that include all positive examples of a concept while also excluding all negative examples is nearly impossible

2. definitional structure implies that category membership, whether something counts as an example of a concept, should always be clear cut

3. definitions appear to be out of step with how people actually use concepts (people do not see categorical subjecting as all or none)

what was Rosch’s study

• challenged the classical view by showing systematic variations in how people rated the goodness of category membership

• bad examples of a category (ie. tomato as a fruit) is often classified differently by the same person across time

• good examples of a category are classified consistently by the same person across time 

• this suggests people are not using definitions for categories at all

what is the modern view of concepts

• the modern view of concepts is that concepts cannot be understood in terms of necessary features but as a prototype

• category membership is based on the degree of overlap with the prototype rather than meeting definitional criteria (no essential features)

• this view naturally accommodates variation in concept goodness and explain why people waiver on borderline concepts

what is a prototype

• a summary representation that lists characteristic features of a category with some features being more important than others

• this expands features that we could consider

what is the utility concept

• implies that we ought to make decisions in a certain way

• for many decisions, we can imagine a singular best choice that confers the maximum benefit

what is the subjective expected utility concept

• we ought to be making choices by considering:

1. the possible consequences of our decisions

2. the probabilities of those outcomes

3. the value we attach to each of those outcomes

• this decision calculus is assumed to apply to all manner of decisions

• we sum the weighted values across attributes to determine the subjective utility of each option

what is the limitation of the subjective expected utility concept

• rational decision making under this framework appears effortful and does not align with how we actually make decisions

what are heuristics

• heuristics are simple rules of thumb that are used to make choices → they can form part of a procedure to simplify the decision process, they can be used to lead directly to a decision, but that decision may not be optimal (they are suboptimal)

• much of our decision making does not require exhaustive sampling and integration of information

• heuristics influence processing of information (ie. descending and ascending numbers experiment)

what is the elimination by aspects heuristic

• where we eliminate individual aspects one at a time which reduces the number of options available

• attribute information is not combined, minimising processing demands

• the pool of viable options is rapidly reduced to a manageable number 

what is the gaze heuristic

• works in order to catch a falling object → where one fixes their gaze on the object and being running towards it and you adjust running speed to angle

• only a single alternate piece of information is used (also seen in the US airways flight 1549

what is the recognition heuristic 

the recognition heuristic (harnessing ignorance) → shown through the German city population test → in situations where we have very little information, we may not be able to assign utilities and weightings to attributes (we can still make non-random choices if only one option it recognised we often choose that one) → this heuristic can even promote accurate decision-making

what is the representativeness heuristic

decisions are sometimes based on the degree to which something (ie. an event, a person) matches a cognitive prototype (ie. the coin experiment)

conjunction fallacy (bank teller and feminist experiment) and base rate neglect (70 lawyers and 30 engineers experiment) are examples of representativeness heuristic.

what is the primacy heuristic

the halo and horn effects in impression formation illustrate anchoring and adjustment (or the primary heuristic) → enduring impressions are influenced by first encounters → initially presented labels can be sticky and influence how subsequent information is processed

what is the framing effect

• the framing effect → decisions are dependent on how the problem is presented to us → when presented with information framed in a postie form of light we tend to pick that option over the negatively framed option

• becoming attuned to how information is framed is essential for making informed choices

what is family resemblance

• examples of a concept tend to shared a family resemblance → they tend to share many features and the specific features that overlap will differ across examples

• family resemblance explains gradations in typicality → typical category members will tend to share a greater proportion of features with other members of the same category and typical category members will tend not to share many features with members of other categories.

what are ad hoc categories 

ad hoc categories share the same structural properties as other concepts (ie. typicality and family resemblance)

what are some challenges to the prototype theory

1. handling category exceptions is a challenge for this theory → ie. the bat is much more similar to a bird prototype than a mammal prototype

2. being highly similar to another example is more influential than being similar to multiple exemplars (Shepard’s exponential generalisation law)

what is the exemplar theory 

• concepts are not summary representations as in prototype theory → instead they are simply memorised instances of a category (ie. category exemplars)

• category membership is determined by overall similarity to all exemplars for one category vs exemplars for other categories

• typicality and family resemblance are explained by typical stimuli being highly similar to many other category exemplars

what are the three core concepts to the exemplar theory

1. similarity

2. hierarchal structure

3. the basic level 

explain similarity

similarity if fundamental to both prototype and exemplar theories → it is often conceptualised geometrically as distance in a psychological space → the closer things are in psychological space, the more similar they are.

what is Tversky’s triangle inequality variations

• according to Tversky’s triangle inequality variations, if similarity can be understood in geometric terms, it must satisfy the triangle inequality:

1. distance (A,B) + distance (B,C) > distance (A,C)di

• if different contexts determine similarity for different stimuli, then the triangle inequality does not apply

explain hierarchal structure

• concepts seem to be apart of hierarchical structures (ie. animals (superordinate categories)→ dogs (subordinate categories))

• hierarchical structure supports inference → features at a higher level in the hierarchy are present at lower levels of the hierarchy

• it also interacts with typicality → people judge features to be more likely to be inherited by typical vs atypical exemplars and for properties to be more likely to generalise from a typical exemplar to other exemplars from an atypical exemplar

explain the basic level

• people tend to use categories at an intermediate level in the conceptual hierarchy → between superordinate and subordinate

• it is informative enough to let us infer useful properties of an object and distinctive enough to avoid confusion with objects from different categories at the same level

• it is influence by culture and expertise (ie. different types of tree experts structure conceptual knowledge about trees quite differently to the average person)