Generalization: effects of learning to spread

Generalization Types

• Vicarious - across people

• Maintenance - across time

• Response - across behavior *

• Stimuli - across situations *

Generalization Gradient: learned reaction can be similar/different depending how close it is to the initial reaction; “close enough”

• Flat Generalization- no generalization, high discrimination

• Broad Generalization- some generalization, some discrimination

• Narrow Generalization- high generalization, no discrimination

How to increase generalization:

• Train in different settings

• give many examples

• different consequences *

• reinforce generalizations *

Discrimination: effects being spread to some situations but not others

Discrimination Training:

• Classical approach: CS+ and US paired, but CS- stands alone

• Operant approach: SD (discriminative stimuli) reinforces and S^ (consequence) does not *

• Simultaneous Discrimination Training: Discriminative stimuli and consequence given at the same time , but stimuli has reinforcing consequence and consequence does not *

• Successive Discrimination Training: SD and S^ individually and randomly *

• Matching to Sample: picking a similar thing out of a list of other things

• Mismatching/Oddity: Picking something different out of a list of options

• Errorless Discrimination: guiding the learner so theres no mistakes

• Differential Outcomes Effect (DOE): multiple behaviors at the same time. Reward one behavior right away and then delay the other reward for a different action. *

Stimulus Control: when training helps behavior respond to specific cues

• Associative Relationship

Concept: any class of characters that have one or more of the same defining features *

Forgetting: losing skill over a period of time with no practice

Forgetting and Stimulus Control: *

• All behavior is influenced by stimulus control (with or without environmental stimuli)

• Forgetting may occur when stimulus control changes due to different environment where learning happened

How to measure Forgetting

• Free recall: repeating information when given the chance to determine hoe much is forgotten, but does not account for partial information

• Prompted recall: repeating information when told to, but not all details are recalled

• Relearning/Saving Method: assessing how much training is needed to get back to initial level of skill *

• Recognition: recalling information without hints and only incorrect/correct answers *

• Extinction method: put a behavior on extinction after retention interval*

• Gradient Degradation: Increased generalization and Discrimination yield higher rates of forgetting

Sources of Forgetting:

• Degree of learning: how much you already know. The more you know, the longer it takes to forget*

• Prior Learning: what you already know. Meaningful content is easier to learn, but it can interfere with recall. *

• Subsequent learning: we forget after falling asleep instead of activities *

  • Retroactive Interference: learning new material increases likelihood of forgetting old material *

• Changes in Context: changing environment from initial learning place can degrade what you remember

  • Cue dependent learning: decrease in performance when stimuli of initial learning period is not present

How to decrease forgetting:

• Over learning: training past mastery to reduce probability of forgetting

• Feedback training: getting positive negative feedback, with positive continuing the behavior and negative reducing it

• Test yourself: period testing yields greater retention than studying

• Mnemonics: device used for active recall (ROY G BIV) *

• Context clues: learning in different environments yield greater retention of new skills *

Learning is not inherited: Automatic responses (reflexes) are natural and do not require learning. Individual learning helps navigate environments and be creative.

Learning and genetics: genetics are important, but environmental stimuli is key

Critical period: period where person learns best and retains the most amount of information

• Harlows Monkey Experiment: monkey prefers comfort and warmth from mother, but lack social skills because that cannot/has not been taught by the mother

Preparedness and Learning: learning is different in different situations *

• Autoshaping: innate tendency to invest in situations that involve food with no need for other reinforcement

Learning Occurs on a Continuum of Preparedness: some things are learned with ease*