Textbook Notes (363,472)
Canada (158,372)
History (100)
HIS282Y1 (1)

law of psychology .docx

5 Pages
Unlock Document

University of Toronto St. George
Denis Walsh

Lecture 3 - Laws of Association 09.29.2012 Temporal Contiguity - Many types of conditioning, some are more meaningful than other ones - Cs – light or tone, US: reward or shock, = delayed conditioned - Types of Conditioning 1. Delay: US is delayed after the CS 2. Trace – bigger delay – could be from context (the US comes to condition, with a trace of the original stimulus, the CS) example - a bell begins ringing and ends just before the food is presented. 3. Simultaneous – both CS and US superimosed 4. Backward – US first then CS (least effective) Nature of US and CS - Conditioning occurs most rapidly when the CS and US are NEW/novel to the subject - Prior exposure to the stimuli can interfere with learning – Latent/retarded inhibition – animal models for schizophrenia (stimulus is not very relevant so you don’t pay attention) - Intensity of CS and US also affect condition – MORE INTENSE = BETTER, not TOO MUCH! Similarity - Some stimuli are more readily associated w/ one another than others?  i.e. tastes and illness? (good) Tastes ad sounds? (not really) Smell and shock? (no) Sounds and shock? (yes) - Internal stimuli go together, external stimuli go together Casual Relevance - Animals attribute a particular effect (fear, illness) with its most probably cause, based on past experience/evolutionary history  Much easier to establish aversion to bitter or our tasting substances to sweet ones  Monkeys and fear of snakes vs. fear of flowers Contingency - Learning requires more than coincidence of CS and US, it requires that the CS predicts the occurrence of the US Delayed is the best ? time is - Rescorla (1968) trained gorups of rats using conditioned suppression of lever pressing. Each group received different combo of probaility of schock during CS (P/US/CS) and probability of shock during absence of CS (P/US/noCS) Stronger association of CS and US = higher rate of learning Human Contingency Judgments - i.e. Financial: Stocks a,b,c,d all go up, is the whole market going up?  Like CS  UIS: 1 stock  whole market  People learn to predict events form experience of relationship Surprise - Leon Kamin’s blocking (1969)  Tone  Shock (phase 1; compound stimulus (tone + light)  shock. Test would be light? What have the new animals learned about the stimulus  Result: NO Learning occurs to Stimulus C in Block group. Why? Stimulus A fully predicts the US so US is no longer surprising and no longer induces learning  US is fully predictable, c becomes irrelevant  Different from second-ordering conditioning because in phase 2, it’s stimulus to stimulus, not stimulus to response Rescorla-Wagner Model (1972) - 1 of the most popular mathematical model of Pavlovian learning that explains a number of phenomena - Principle tenet of model is that learning occurs on a conditioning trial only is the US is surprising (violation of expectation/ error-driven model -“The more your expect something to happen, the less the learning “ - Effects of US magnitude on λ – the bigger the US, the higher the asymptote - Another twist on the equation:  Rescorla and Wagner proposed that the degree to which the US is predicted on a trial depends, not only on a signal Cs, but on ALL the CSs that are present on that trial  If 2 stimuli are presented in a compound together, then the total associative strength of the compound stimuli is the SUM of the strengths of the elements Blocking - Training Stimulus A to the asympotitic level - How would you learn anything about stimulus B though? - If we predict with R-W equation, we’d assume no learning Unblocking - What can you do to overcome the blocking? Increase US magnitude in Phase 2 - If we increase US magnitude, we increase λ, and learning can now begin - Model predicts increase in associative strength to Stimulus B in first trial of Phase 2 Rescorla Wagner Model - With (-) associative strength, tests that would work would be summation dn retardation. Why? - Quantification of surprise is the basis of the model - Model effectively accounts for the fact that conditioning always involves compound stimuli (context, discrete cutes) - US expectancy is important in the learning of CS-US associations - But there are some phenomena that are not explained by the model (i.e. latent inhibition) Example Group 1, Phase 1: Stim A along, Phase Stim A – US – Result? Group 2 NOTHING; Phase 2: Stim A – US– Result? What would r-w predict, Group 1: No US so Va= 0 (group 1) Group 2: Use equation, α is 0.3, λ is 1, Va = 0, = 03. Actual r
More Less

Related notes for HIS282Y1

Log In


Don't have an account?

Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.