Module 3 Study Guide.pdf

Module 3 Study Guide Classical Learning I. Begins with a reflex (UR) to a reflexive stimulus (US). A. When a neutral stimulus is paired with the US, it becomes a learned stimulus (CS) which produces a learned response (CR). B. Example: Pavlov’s dogs...

Module 3 Study Guide Classical Learning I. Begins with a reflex (UR) to a reflexive stimulus (US). A. When a neutral stimulus is paired with the US, it becomes a learned stimulus (CS) which produces a learned response (CR). B. Example: Pavlov’s dogs 1. US (reflexive stimulus): food 2. UR: salivation 3. CS (learned stimulus): bell 4. CR: salivation II. Common conditioning paradigms A. Conditioned emotional response 1. Why? a) Already have Skinner boxes b) Uses an emotional reflex c) Rapid conditioning 2. US: shock to floor grid, freezing 3. CS: most neutral stimuli 4. CR: suppression of ongoing behavior B. Example: child hearing sound of thunder 1. US: sound of thunder 2. UR: startle reaction, fear 3. CS: seeing sky get dark 4. CR: seeing dark clouds, then crying C. Eye blink 1. Uses a monophasic reflex 2. Slow conditioning 3. Easy to measure size of CR a) US - airpuff b) UR - eyeblink c) CS - picture of air balloon in dilation machines d) CR - eye blink D. Galvanic skin response 1. Human participants 2. Not under conscious control a) US - mild electric stimulus b) UR - defensive arousal c) CS - any neutral stimulus d) CR - increased electrical resistance III. Laws of Learning A. Acquisition - deceleration of learning B. Classical learning is different from habituation 1. Habituation - only 1 stimulus (reflexive stimulus) 2. Classical learning - 2 stimuli (US and CS) 3. Relationship between US and CS → law C. Extinction 1. Spontaneous recovery - extinguished CR reappears after a delay 2. Rapid relearning + extinction → two types of relearning 3. CS and US → reacquisition 4. No US and CS → rapid reextinction D. Overlearning 1. If learning continues even after there is no longer any visible change in behavior → spontaneous recovery shows the effects of this overlearning 2. Example: a) Ring bell but don’t give dog food b) Dog stops salivating → takes break c) Less spontaneous recovery E. Generalization 1. Habituation to one stimulus generates to similar stimuli a) Ex. trained hunting dog may react the same way to fireworks as it does to rifle shot 2. Generalized response weaker than CR, stronger when general stimulus is more similar to CS 3. Do NOT pair new stimulus with US! a) Want - make old response to new stimulus b) Not - make new response to new stimulus 4. Important because a) Stimuli vary in specifics but have the same significance F. Discrimination 1. If two similar stimuli have different significances → we learn to discriminate IV. Early Classical Conditioning A. Stimulus strength: strength of US 1. Higher asymptote 2. Faster learning B. Stimulus strength: strength of CS 1. Faster learning C. It is possible to get more CR when CS is absent than when it is present D. Pavlov’s Principles  1. Stimulus substitution a) Part of brain excited by US b) Now CS is activated → we build connections c) Created path from CS to US → CR 2. Co-occurrence a) US produces UR → CR needs to be presented 3. Equipotentiality a) Any neutral stimulus is equally good 4. All three principles are WRONG V. More Complex Antecedents A. Many things present at same time as potential learned stimuli (CS) B. Multiple CS → no problem! C. Higher order 1. Play a video game (learned stim) + release of endorphins (reflex stim) → enjoyment (reflex response) 2. Over time: game by itself → enjoyment 3. Friend comes over to play → enjoyment 4. Friend in other settings → enjoyment 5. Continues so on and so forth D. Blocking 1. What if a new stimulus appears when another CS is established? a) Pavlov predicts learning to new stim → co-occurrence → learning b) No conditioned response to second stimulus VI. Theories of Classical Learning A. Why blocking? 1. Learning is not based on co-occurrence 2. Based on surprise a) Learning is driven by unexpected events 3. ΔXᵢ = Sᵢ ✕ (U - XΣ) a) Xᵢ = possibility that US will occur b) Example: stimulus, tone and light (1) XT = expectation of US given tone (2) XL = expectation of US given light (3) U = intensity of US (usually set to 100) (a) Set to 0 when there is no US (4) XΣ = overall expectation of US given all of the CS (a) XΣ = XT + XL (5) Sᵢ is importance of the CS (6) U - XΣ = “surprise term” 4. ΔXT = ST ✕ (U - XΣ)  a) ST = 0.1 is importance and determines rate of learning b) US is presented so U = 100 c) As trials continue (1) Expectation builds (2) Surprise lessens (3) Learning continues at a reduced rate 5. Basic explanations a) Deceleration of learning b) Extinction (1) Start by expecting possibility of US (starting X = 100) and no intensity of US (U = 0) c) Higher asymptote with more intense US d) Faster learning with more intense CS e) Spontaneous recovery f) Blocking (1) Surprise is based on expectation given both stimuli VII. Attentional Theories of Classical Learning A. Limited attention B. Learn to pay attention to stim that co-occur within US C. Blocking: 1. Already have full attention on tone so ignore the light → no learning to light D. CS pre-exposure effect: 1. Present potential CS (tone) alone over and over initially 2. Now pair tone with food 3. Learning of tone is slower E. Most current theories include both expectation and attention components, no current theory explains everything F. Informative value of CS 1. A clever CER experiment a) Random presentation of shocks (US) b) Vary probability of tone (CS) during shock c) Co-occurrence: CS + US = 20% (1) Extremely slow learning d) Informative value: shock more likely during tone 20% than no tone (1) Tone is CS+ e) CS - US relationship can be probabilistic 2. S-S: associate one stim with another 3. S-R: associate stim with response  a) When a dog hears a bell, does he think “food’s coming” or “time to drool”? 4. If US does not produce UR a) S-S: no CR b) S-R: yes CR 5. S-S: dog learns bell → food → drool a) When dog is full and hears bell → does not care about the food 6. S-R: full dog hears bell and still thinks about drooling 7. Standard CER with loud noise a) Stage 1: condition CR to tone b) Stage 2: habituate to US so now no UR c) Stage 3: test CS (1) CR? (a) S-S: No! (b) S-R: Yes 8. Higher order conditioning a) Stage 1: condition CR to tone b) Stage 2: 2nd order CR to light c) Stage 3: habituate to US so now no UR d) Stage 4: Test CS(L) e) S-R learning in HOC 9. CS organization a) If you see a flash and hear a bang at same time → may be part of same event b) How to tell (1) Condition light and tone at same time (2) Extinguish response to tone (3) Test for response to light (4) If response to light not extinguished → light and tone are 2 different stim (5) If response to light also extinguished → light and tone are single stimuli G. Gestalt laws 1. Law of similarity - similar things grouped together 2. Law of temporal proximity - things close together in time are more likely to be grouped 3. Law of spatial proximity - things close together in space are more likely to be grouped H. Biological considerations 1. Homeostasis - physiological systems have a preferred state  2. Habituation - keeps system on an even keel a) Preparation - classical learning acts like habituation b) Learned response - keeps system on a more even keel 3. CR alone produces only compensatory rebound effect a) US - moderate shock b) UR - increased heart rate c) CS - beep d) Test CS alone → CR is decreased heart rate 4. Pavlov wrong → CR ≄ UR 5. Conditioned taste aversion a) US - gastro-intestinal illness b) UR - regurgitation c) CS - food item d) CR - avoid food (1) Brief CS-US delay (a) Strongest after delays in hours (2) CR is same as UR (a) Not similar (3) All potential CS equally good (a) Taste CS much more likely than other senses I. Final notes 1. Building an established reflex makes it simple 2. Laws of learning have evolved that make classical learning adaptive and flexible 3. Learning evolved through variation and selection through generations