Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
134 Cards in this Set
- Front
- Back
Learning theory is a comparative science which means |
it tends to rely on comparisons between aniimals and humans |
|
In learning theory we are referring to more than just |
Learning at school. It refers to learning in general |
|
Learning is defined as |
an enduring change in an organism from past experience, including past events and practise. which leads to a change in behaviour i.e. you need to be able to reproduce it later |
|
Learning is not the same as |
Performance. Performance is affected by learning but also depends on you abilities, your motivations and opportunities to show abilities e.g. the difficulty of test, unfairly hard test = can't show what learnt, and not perform well but not coz didnt learn |
|
A reflex is |
an innate change in behaviour simple and automatic Serve a general function e.g. rooder reflex (light pressure on baby cheek = tries to suck on it (evolutionary to find food) |
|
Instinct |
Series of behaviours that are genetically determined (i.e. typical of all in species). No learning required to do this. e.g. bird ritual dance - clear stage despite no leaves (do it just coz its a instinct, not necc with end goal). |
|
Maturation is |
Aging. mostly physical changes but also affect learning over lifetime e.g. learning to stand->walk->run |
|
Fatigue |
Change in behaviour for this is unstable Its a physical state of discomfort and lose efficiency coz of tired, bored, no rest, stress etc. May result in can't do something that has been learnt |
|
Reflexes can even be seen in unborn children/young kids
|
e.g. knee jerk, pain -> Withdraw (pain = elicit stim, and withdraw is the correspoinding response) Momo Reflex: ie complex startle reflex, from loss of support/feeling of falling. The babies then spread their arms and legs out, and try and grab on, and call out to mum |
|
Reflex arc is |
the fastest path via neurons that lets us do the reflex as quick as possible. Sensory stim --> sensor --> spinal chord --> motoer nerves which stimulate muscles Weaker stim = less muscle response. |
|
Why we use animals for studying learning |
Cheaper Wider Scope of study that can be conducted (regulated by ethics) animals chosen would be most similar to human according to what's studied. |
|
Habituation shows |
lower response to a stimuli because of more exposure. Stimulus specific ie will only get bored of the one stimulus |
|
Habituation is not the same as fatigue or sensory adaptation |
Fatigue adapt = muscle incapacitate, cant respond Sensory Adaptation= temporary insenstive to stimulus |
|
Sensitisation is |
More response becaus of more exposure. e.g. drugs: more cocaine in rats = runs more |
|
We use habituation and sensitisation to work out |
what needs to be ignored and what needs to be attended to: ie sift through all of the sensory info around us |
|
Classical conditioning: |
Pavlov, 1927, Bell/Light (CS) KEY: the conditioning is out of your control |
|
Pavlov said theres a psychic reflex |
ie what he called a conditioned response, which gave the name "conditioning" |
|
US UR CS CR-> |
i know these yay
|
|
Second order conditioning |
ie conditioned stim gets paired with the US enough that the conditioned response gets transfered to the CS, and can act like a US itself Using this can transfer CR from one CS to another CS |
|
Acquisition and Extinction |
Acquisition is when you pick up the pairing between CS and US Extinction is when the pairing gradually phases out (e.g. show the CS alone without the US = decreased CR) |
|
Strength of Conditioned reinforcers can be measured |
e.g. Pavlov measured how much saliva came out |
|
In human psychology, classical conditioning helps us study... |
Our learning process |
|
In male animals, a strong US is |
the opportunity to even see a female of same species (SEX SELLS) Also works in human (pair product with what you think they want) -> view is either oblivious or doesn't care |
|
CRis not necc the same as UR |
e.g. Fear Conditining: Rat in box, signal = shock. |
|
Extinction isnt just unlearning |
There is such a thing as spontaneous recover i.e. after you let the pairing extinct for a few days, when you test the CS again later, will still get the CR |
|
Examples of Classical conditoing |
Food Preference(neutral tast pair with sweet) Conditioned fear (little albert) |
|
Appetitive vs Aversive |
Rewardingvs punishing |
|
Anexample of aversive conditioning |
Whenever you alki and puke, you feel like puking whenever you see that same alki this is called Anticipatory Nausea Nurse-> give chemo shot->feel sick, so start to feel sick when see nurse. |
|
Frequency of pairings affect classical conditioning coz |
the more pairings (US and CS) that you present, then stronger the CR becomes, but it has a ceiling effect i.e. theres a maximum CR strength achieved by more pairings (diminishing returns) |
|
CS Intensity affects classical conditioning via |
Physical salience of CS = more effect. Weaker CS takes longer time to reach max CR. |
|
Salience of US affect classical conditioning |
determines the max learning value (weak US = low max learning) e.g. like one stimulus better = respond better |
|
Contiguity is |
the time between presenting a US and CS Its better to present these at a reasonably quick, optimal interval i.e. not instantly show it, but at the same time don't take a long time to pair it) |
|
ISI stands for |
interstimulus interval - ie a measure for contiguity |
|
Strenght of CR depends on the ISI |
There is an optimal ISI to aid learning, which depends on the task e.g. if the task is easy then quick ISI, if hard then allow appropriate time Exception is conditioned taste aversion, when longer ISI = stronger CR |
|
Contingency |
Is the probability that the US is paired with the CS. and probability the US happens in the first place Positive contingency means higher probability of showing the US negative contingency = no real effect of CS Its important that show CS means US will come to aid learning |
|
Things affecting learning in classical conditioning |
High freq = higher learn more intense US = more learn more intense CS = faster learn closer timing of CS and US pair = better learn higher positive contingency = more learn |
|
Thorndike was the guy who |
tested animal intelligence using a puzzel box: showed that animals use a trial and error approach with improved performance over time, and don't get "eureka moments". From this he made the Law of Effect: In any situation, if you did something and got a positive outcome, in the same situation you will repeat that action |
|
The law of effect implies that |
behaviour is strongly influenced by what has happened in the past.
|
|
Discrete Trial procedure: |
Each time there is a single trial where you need to "reload" the subject e.g. you put a mouse in the maze, it finishes the maze, you put them back at the start |
|
Free Operant procedure |
ie use something like a skinner box, where ypu put the rat inside, and they do an action, get some form of reinforcement (food, water) |
|
Use of the skinner box was big part of radical behaviourism, which believed |
that the only things that matter are the behaviours we can see, and didn't really care about whats going on inside the mind Reaction to psycho analysis influential from 1900-1950s in USA |
|
Operant Conditioning |
is different to classical conditioning because it requires action from the subject |
|
Reinforcers are |
things that make a subject do an action MORE |
|
Punishment is a |
thing that makes a subject do an action LESS |
|
Intrinsic value of reinforcers is important to consider, |
e.g. use of primary reinforcers i.e something the subject will definately like
Secondary reinforcers reinforce via experience, ie bell showing food coming/ dog clicker Social reinforcers are intangible but can also shape behaviour e.g. praise and attention |
|
To make the subject do what we want them to, we use shaping. Shaping is |
When you reinforce a subjects actions the closer it gets to what you want them to do, and when you make the criteria to get reinforcement more strict as they get closer |
|
using shaping, it is possible to do behaviour chaining, which is |
When you shape a bunch of simple behaviours in a chain, like a combo. |
|
Shaping can occur just from environment i.e. it wont always form a desirable outcome |
An example is when kids throw a tantrum in public, and get rewarded, then they start to throw more tantrums in public Also makes superstitious behaviour e.g. win in gambling when wearing lucky shirt = wear it to casino all the time. Also pigeons displayed superstitious behaviour when looking over head and rewarded. [CHECK LEC RECORDING 3] |
|
4 types of instrumental learning are |
Positive reinforcement: give something nice
Negative reinforcement: take away something bad Positive punishment: give them a punishment Negative punishment: Take away nice things |
|
Escape and avoidance learning is linked to |
negative reinforcement. By successfully avoiding the shocks, the comfort is reinforcing, so they learn to escape. This also makes phobias hard to treat (ppl naturally want to avoid, but to treat need expose) |
|
Reinforcement schedules affect instrumental learning. Schedules include |
ratio vs interval (i.e. no. of trials vs time) Fixed vs variable (i.e. always after X trials vs on average, after X trials) Contiunous vs partial (all the time vs sometimes) |
|
Using a variable schedule will lead to a mindset where |
More action, means that I am more likely to get reinforcement (as opposed to fixed where the pattern is obvious) |
|
The instrumental reinforcement learning schedule would be |
Fixed schedules (and notably fixed ratio is the fastest method) |
|
Which would lead to better learning: A very high quality, pleasing reinforcer OR A very predictable reinforcer |
The really predictable one is better |
|
The cycle of drug abuse starts from |
positive reinforcement from the high. Next you get the negative reinforcement of hangover by taking more drugs |
|
In other words instrumental learning affects behaviour according to whether |
the reinforcement/consequence was appettitive or aversive, and whether it was an action or lack of action that bought about the consequence |
|
Stimulus Control is when [CHECK LEC RECORDING4] |
In conditioning, an animal's actions are affected by learning that there is a difference between stimuli. These stimuli are called the Discriminative Stimuli [TLDR different actions are controlled by different stimuli that they are paired with] |
|
Skinners ABC Threesome: |
Antecedent: Stimulus controlling the behav Behaviour (response being reinforced in sitch) Consequence: whether positive or negative Discriminative stim is part of Antecedent |
|
Normally the process for stimulus control is |
Sd (discriminative stimulus) --> Response --> Sr (reinforcing stim or punisher) |
|
Learning related factors which are important in stimulus control include |
Basic conditions neccessary which are important classical and operant conditioning |
|
Performance related factors for stimulus control include |
how the behaviours from one situation get transferred to a similar, new sitch. Also effects generalisation/discrimination |
|
Generalisation refers to |
how much a behaviour transfers to a new stimulus |
|
Discrimintation refers to |
how much a behaviour doesn't transfer to new stimuli |
|
Watson and Little Albert was an example of |
being conditioned to fear a rat, and the CR (fear of rats) was transfered to other similar stuff |
|
Rachman:
in experienments about acquisition of Fetish via classical conditioning |
took people with no fetish and paired CS (high leg boots) with porn (US)
Measured erections Found gradual acquisition, and extinction over a week. However, after the week of extinction they exhibited spontaneous recovery |
|
Another example of generalisation |
Pigeons with colour: if you pair 1 colour with food, they will peck a lot if its same colour/similar but the less similar the color becomes, the less response/pecking from pigeon |
|
Marks and Gelder 1957: |
ie paired the fetish with shocks (measured the reaction time for CR, longer time to respone means fetish was reduced?) |
|
Discrimination learning can be taught through |
training and different reinforcement schedules. i.e. if you reinforce one, and not another, then subject can offer tell it apart. |
|
Classical and instrumental examples of discrimination learning: |
Instrumental example: high and low pitch is presented with action for high tone = rft and low = no rft. i.e. high tone is do action, but not for low] In Classical high tones means rft is coming, while low means no rft coming |
|
Discrimination learning |
S+ and S- By having 2 , one positive and one negative attached to different stimulus, the subject gets cued to ignore one and to attend to one (the positive rft one) |
|
For complex discrimination |
reinforce the "correct" decision against other "incorrect" choice by same size, shape, no, textures etc. N.B for this to work they need to be able to tell the difference between these. |
|
Animals can actually learn to discriminate specific stimuli accurately, e.g. |
pegions and other animals can learn categories and high level discriminations through discrimination learning e.g. Pigeon can be trained to identify stylistic similarities e.g. in Art: if pair monets with rft, vs other art, then they can learn to identify the style at better than chance |
|
Generalisation in humans occurs based on |
physical attributes, similarity in meaning for words, and rules/analogies that link things that wouldn't otherwise seem linked. N.B we generalise in all types of ways, no one set way |
|
Social learning is a |
highly comparative process |
|
As opposed to instrumental and classical conditioning where you have to directly experience it, social learning |
is from watching models i.e. acquire behaviour by watching others |
|
Behaviourist said that behaviuor is sourced from conditioning and generalisation (own experience) but... |
This doesnt account for social conditioning, ie acquiring behaviour from watching others |
|
An example of social transmission: |
Birds throughout a whole country opening milk bottles, which: SD - milkman leaves Resp: peck lid Rft: cream OR Cultural transmission of preferences |
|
Social facilitation vs Social learning |
Social facilitation is when a social being is in a social situation, and it helps their learning, not necc social learning. examples are: Goal enhancement: i.e. achieve goal, make more accessible for mates (e.g. poke milk lid & so others can have it too) |
|
Observational Learning is when |
There is an observer (lab monkey) and a performer (wild monkey)
the lab monkey has never seen snakes, so doesn't react to it initially BUT once they see the wild monkey react, they learn this behaviour. It is also generalised to all snakelike things US = wild monkey UR /CR = fear of snake CS = Snakes |
|
Biological preparedness to learn some behaviours was displayed when |
Monkeys learned to fear snakes and related things, BUT when showed an edited video of monkeys scared of flowers, didn't change behaviour. (i.e. through evolution specialise to not fear some) |
|
The two action test used to see if animals can pick up instrumental actions from observing peers (seeing consequences of their actions). Evidence of this was when |
A quail watch 2 other quails, and copies the one that got rewarded |
|
In the experiment where the ape and kids was given a tube, and had to open the tube... |
They were divided into 4 conditions: 2) is shown just that there is a reward in there. 3) is shown just an ape with the tube. 4) is shown the action of opening the tube only 80% of the time get the treat, but only 20% of apes did exactly same method, but most did it their own way. Kids were diff: a bit more copied, and performed better if they saw the action. i.e. chimps tend to freestyle (emulation) and kids tend to imitate |
|
Types of social instrumental learning |
mimicry: copy regardless of goal (blind copy) emulation: achieve but freestyle the metho imitation: copy with same method of performer, and know the goal ie motivated from other's reward (i.e. maybe after emulate, you still can't so you decide to imitate) |
|
When Bandura tested the kids, he did two tests. which tests did they do? |
they did 2 tests because want to test memory The first one was rewarded to test memory The second one was normal i.e. just see what they do after seeing the video (there were 3 video condition: violent = reward or punish, and also just acti normal) |
|
Modelling: |
Children imitate specific adult behaviour, but can also model generl styles of behaviour. e.g. aggressive vs gentle play |
|
The suggested process of how social learning works is |
1) watch and learn how they do it 2) use info whre useful 3) info not necc used instantly |
|
Bandura:see how reinforcement affects modelling |
when model is rewarded, more likely to learn what they did When model is punished, less likely to learn what they did Bandura's main finding was that this can happen via TV (they used to think violence on TV was calming, but it actually can be modelled by kids |
|
Social cognition theory |
Says theres a few steps to learning e.g. in Bobo dolls. 3) assess if you have ability to copy 4)be motivated to copy e.g. if they got reward |
|
Applications of social learning in |
Ad campaigns: via response/rft association e.g. anti smoking Parenting: smacking violent child only displays that violence is the way to solve things |
|
Motivation is |
The reason that we do things it is a neccessary, temporary state that enerises behaviour Temporary implies that it changes over time |
|
Hebbs Analogy of Motivation: |
In a car the engine provides power, the steering provides direction In BEHAVIOUR, MOTIVATION provides the power and our LEARNING determines how we choose to behave and interact with our enviro |
|
Lorenz, Timbergen were ethologists who found |
Fixed Action Patterns. Some examples include Sign stimulus: i.e. some stimulus that activates an FAP. e.g. for sticklebacks its the color red Supernormal Stimulus: i.e. their evolutionary tendencies tend to make the animal react to a exaggerated, (and sometimes desirable) trait e.g. goose chose big egg, so choose football |
|
Instinct can also be described as being |
very stereotyped (the actions are mostly same, but can vary slightly from the animals personal experience) |
|
the purpose of the stickle back experiments (where they compared a stickle back in mating season's reaction to a normal looking stickleback to progressively less stickle back like stimulus, but with red bellies) |
was to see which one is causing the FAP: was it the male or was it the red? (Answer: It was the red.) |
|
Fixed Action patters are characterised as |
Standardised behaviour for all species toward same stimulus Is not a reflex but a combo of behaviours (complex, and once started the entire combo has to be finished) Regulated by biological state or stimuli in the environment |
|
In terms of the process for FAPs to work, |
1) the animal starts in a biological state A (e.g. damn is it october? I'm horny AF). 2) it gets an input from the environment (e.g. sign stim -> male stickleback = that full belly tho!) 3) Input activates a neural instruction to start FAP 4) BEGIN THE FAP(ing) |
|
Fixed Action Paterns are not directly |
motivated towards an end goal. They are simply boought out by enviro and bio circumstances |
|
There are no clear FAPs in Humans, though maybe some instincts/reflexes. Examples of proof used for seeing instincts etc in Humans include: |
Biological Basis : gene sequencing (no evidence found yet) Cross species similar: e.g. similarites in behav between humans and great apes Crossculture similarities Twin Studies: study similarity of behaviour between identical twins seperate at birth (rare studies obviously) Development studies: e.g. reflex in newborns |
|
Eibi Eibesfeldt is the guy who |
Went around the world and studied people in their "natural habitat" by filming them with his weird camera without their knowledge. He saw some cross-cultural, nonverbal signs that seemed to be quite common e.g. albert's eyebrow flash, acting coy (flirting) |
|
Problems with instincts |
Circularity: i.e. kangaroos exhibit mobbing instinct because they mob together. But they mob together due to a mobbing instinct Proliferation: i.e. new behav, can't explain... ITS A NEW INSTINCT! |
|
Learning and Motivation can be linked together through |
Drive theories. |
|
Specific Drive theories specify that |
A Drive sensitises you to something that may reduce that drive It then motivates you to go and interact to reduce drive |
|
Specific drives have some issues such as |
Circularity |
|
Hull made |
General drive theories |
|
General Drive theories say that |
we all suffer deprivations, which create a need. |
|
Process of General drive theory says: |
1) We feel a need (e.g. hunger) 2) we feel a tension as a result - a Drive 3) We then do random stuff to try and reduce the drive (e.g. fidget, forage for food) 4) During random activity, you find food, and reduce the drive. At the same time, you notice things in your surroundings. From this, you learn that that thing in your surrounds can help you reduce the drive. (reduction of drive helps you to create a link between that thing and reduce drive) |
|
Behaviour strength is found by |
Habit x Drive i.e. how strongly you have learnt the behaviours x how motivated you are to do it e.g. hungry rat vs not hungry rat AND trained hungry rat (push lever= food) vs oblivious hungry rat (trained rat =push lever harder) |
|
Overall the drive theories are an |
Sr model, i.e. you see a stimulus, and it makes you do something in response (automatically) but this is OVER SIMPLIFIED |
|
Advantages of General Drive Theory over Specific drive theory |
While Specific drive theory has circularity and homunculus problem, General drive theory is general and doesnt "prescribe"/dictate a behaviour in a certain situation, instead it depends on what the animal learns. |
|
Homeostatic Needs vs Non Homeostatic needs: |
Hoeostatic is related to survival e.g. eating. Sex drive = biological need for genetic success i.e. survival of species, not self. |
|
Approximal motivation is |
related to needs that deal with your own survival (i.e. needs close to yourself)
|
|
Distal motivation is related to |
needs for survival of a species (i.e. reproduction, needs that are far from self and related to the entire species) |
|
Problems with drive theories |
1) Drive reduction is not needed for reinforcement (e.g. sugar vs saccharin, both are reinforcing, but saccarin doesn't make you feel more full) 2) Stimulating a drive (but not satisfying) can be reinforcing e.g. pulling in Brave frontier 3) Ignores whether you like/prefer some reinforcers over others |
|
Examples of drive reduction not being the only motivator of behaviour |
Harlow & wire mom: baby monkeys still prefer the cloth mum despite wire mum reducing its drives Sensory Deprivation: i.e. even though unrelated to reducing a drive, we can't stay in sensory deprivation (almost like a moderate level of sensory stim is pleasing) |
|
Incentive Value of a stimulus is determined by |
How much you like it How much you biologically need it Your Current arousal |
|
Humanist approach emphasised that |
We are different from other species because we have freedom of choice and desire to reach our full potential. |
|
Carl Rogers believed that |
I.e. we want to achieve full potential i.e. Fully Functioning human. We also have a self concept (whether negative or positive) but e.g. if we have a positive one and someone breaks it, then we encounter pathology |
|
Carl Rogers believed that to solve mental problems |
To do so we need unconditional positive regard (which means positive intentions & put it nicely, not just compliments)
Conditional positive regard = tension, anxiety etc and start of problems |
|
Roger's fully functioning person has the follwing traits |
open to experience Lives in the moment (existential) Trusts self Free to make choices Is creative and can adapt |
|
Maslow studied mental health rather than mental illness |
and propsed that our needs are ranked in a heirarchy (Maslows Heirarchy), and we scale these needs to become a fully function/Self Actualise person |
|
Thig is that Maslow's heirarchy has a bunch of critisisms |
1) elitism & culturally insensitive (fits mostly USA ppl) 2) Some are weird (mystical exp, political) 3) Seems arbitrary, i.e. just some traits he likes/has 4) Hungry and scared for safety at same time? |
|
Main criticism of Humanistic approach |
Empirically Weak |
|
Old School way of assessing human goals was made by Henry Murray |
Projection tests, e.g Rorschach, Thematic Apperception Test (TAT) i.e. they assume that when provided with ambiguous stimulus, a subject will project their thoughts that they keep thinking about into their interpretation |
|
LT Human needs: |
NAch |
|
A high level of NAch is linked to which career? |
Being an executive N.B casual relationship unclear i.e. unsure if its High NAch people who become executives, or if working in an executive position causes you to become high NAch |
|
If you have a high NAch then you are more likely to
|
take pride in completing a task
|
|
Atkinson and Letwin (1960) found taht |
People with high NAch would predicatably choose a moderate difficulty task, while others chose at random. (though high NAch were expected to choose hard tasks) |
|
The reason that Atkinson and Letwins finding makes sense is |
The Expected utility of success: ie value of success vs probability of success. i.e. to them its very important that they succeed, so a task too hard /easy is not desirable. |
|
McClelland found that for high NAch people, they prefer |
Tasks that involve personal responsibolity in problem solving Moderate tasks Rapid and concrete feedback (find this rewarding) |
|
Legacy of Humanists |
It was a reaction to previously unappealing predominant disciplines (Behaviourism and Psychoanalysis) While conceptually appealling (since they categorised needs etc) but WEAK evidence |
|
Legacy of Atkinson & Letwin, Mclelland and Murray |
Findings made from observation (i.e. a bit more data driven than humanist), from which theory assessed and modified and Good for predicting behaviour (GOOD) |
|
Not really a big point In classical conditioning, a cue in an environment doesn't immediately change behaviour (e.g. maccas sign) |
invokes appetative change (think of the chips = yummy) Change behaviour (not hungry but seek food) |
|
In classical conditioning, when you learn a US CS pairing, they do show a reflexive response (e.g. freezing). |
This is proof that there is some learning going on and its to do with different internal states e.g. motivation (avoid) and emotion (fear). This is not however a transferal of response from one cue to another. |
|
What is Skinner's criticism of Thorndike's discrete trial procedure? |
When the subject can respond is constrained, and there is only one response/reinforcer per trial. Also, there is handling stress. |