Learning Theory

Learning Theory. Rewards and Behavior.

Classical Conditioning. Pairing a neutral stimulus with an unconditioned stimulus to create a conditioned response Pavlov.

Classical Conditioning. Unconditioned stimulus - leads to an unconditioned (automatic) response Neutral stimulus - does not lead to an automatic response Conditioned stimulus - was once a neutral stimulus, but now elicits an automatic response Conditioned response - automatic response to a previously neutral stimulus.

Operant Conditioning. Operant Conditioning - "active behavior that operates upon the environment to generate consequences." - Skinner Law of Effect - actions followed by desirable outcomes are more likely to be repeated, while actions followed by undesirable outcomes are less likely to be repeated - Edward Thorndike Reward - increases behavior Punishment - decreases behavior Positive - adding something Negative - removing something.

Quadrants. Positive Negative Reinforcement Add an appetitive in order to increase behavior Dog gets a treat when they sit Remove an aversive in order to increase behavior Leash pressure is released when dog walks by your side Punishment Add an aversive in order to decrease behavior Dog is hit on the nose when they bark Remove an appetitive in order to decrease behavior Dog does not receive attention while jumping to greet people.

ABC. Antecedent - factors that lead to a behavior Cues Environment Triggers Behavior - the response to an antecedent Consequence - what happens immediately after a behavior.

Humane Hierarchy. Hierarchy of Behavior-Change Procedures Positive Punishment Extinction, Negative Reinforcement and Negative Punishment o.-..a'0 Differential Reinforcement of Altemative Behaviors Positive Reinforcement Antecedent Arrangements Health. Nutrition. and Physical Setting 2013.

Reward Schedules. Fixed - Number of responses or amount of time is set Variable - Number of responses or amount of time changes Ratio - Behavior occurs a number of times before reinforcement Interval - Amount of time passes before reinforcement.

Reward Schedules. Fixed ratio - Reinforcement is given every set number of times a behavior is performed Treat every 3rd time the dog sits Fixed interval - Reinforcement is given every set amount of time Treat every 3 seconds the dog remains in a sit-stay Variable ratio - Reinforcement is given after a random number of responses Sit-treat, sit, sit, sit-treat, sit, sit-treat, sit, sit, sit, sit, sit-treat Variable interval - Reinforcement is given after a random amount of time Reinforce for a 3 second stay, then 10 second stay, then 1 second stay, then 15 second stay.

Reward Schedules. Continuous reinforcement - Reinforcing every time is good for teaching a new behavior Variable - Reinforcing for random intervals or ratios is good for maintaining behavior. It is more resistant to extinction Extinction - A previously reinforced behavior is no longer reinforced.

caninehabit.com FIXED number of times or amount of time that passes is set number of times or amount of time that passes is random an amount of time needs to be passed before reinforcement moments of reinforcement pane m 15 TIME ELAPSED (s) moderate. stea&y response pattem moments of relnforcement 15 TIME ELAPSED (s) behaviour needs to occur a number of times before reinforcement mmyts ot reinforcement post-rel "forcement pause 15 NUMBER OF REPETITIONS hiji, ste*y reverse patt«n roan—Its NUMBER OF REPETITIONS.

Criteria. Difficulty Duration Distance Distraction.

Life Rewards. Go sniff Pinecones/sticks Excited voice, jumping/running together Favorite trick.

Preference Assessments. Free Operant Observations - choose between many different items, ex free play Multiple Stimulus without Replacement - choose between lineup of multiple items, when one is selected it is removed from the options. Can easily determine hierarchy of preferences Multiple Stimulus with Replacement - choose between lineup of multiple items, the selected item is placed back into the lineup. If one is continuously selected above everything else, then you know it is of high value, but gives no information on the comparisons between other items Paired Stimulus - choose between 2 items Single Stimulus - single item presented. Interaction with and duration of engagement are recorded. Unlikely to be used for our purposes https://ebip.vkcsites.org/preference-assessments/.

Homework. Think about strategies and rewards your dog might like. Try out at least one life reward. Conduct a preference test. Rank your dogs preferences. Multiple repetitions over time can give you wider insight or more certainty. If you use both, conduct assessments with food separately from assessments with toys. Be creative! Socks? Cardboard milk carton? TV remote?.