Paradigm Decade1) Theory Key concepts
(Connectionism)2) 1880 - 1900 Connectionism (Thorndike) - learning is incremental strengthening of the S-R3) association
- S-R associations are strengthened through repetition
- outcome of a S-R event can strengthen or weaken the connection
- potential to learn leads to frustration if not satisfied
Behaviorism 1900 - 1910 Classical conditioning (Pavlov) - learning is a visible change in one's behavior
- learning is manifested in a natural reflex reaction on an associated environmental stimulus
- emotional response can also be learned or conditioned
1920 - 1930 Contiguity theory (Guthrie) - behavior is formed by a series of movements which are learned through S-R associations
- a close temporal relationship between S and R is necessary for learning to occur
- learning occurs on first experienced instance of the stimulus
- reinforcements (reward or punishment) do not influence the strength of this connection
Neo-behaviorism 1930 - 1940 Sign learning (Tolman) - suggests studying behavior on the molar level (whole, purposeful, goal-directed behaviors)
- learning is acquisition of knowledge through meaningful behavior, not mechanical moves
- rewards or punishments can only be used as motivators for performance, not learning
- animals are not simple mechanisms, but intelligent organisms capable of cognitive processes
Drive reduction theory (Hull) - mathematical formulas attempting to explain behavior and the likelihood of its appearance
- drive (a stimulus in form of a biological need) results in behavior in order to satisfy it
- reinforced S-R learning through the reduction of a biological drive
- cognitive factors need to be taken into account when explaining human learning
1950 - 1960 Operant conditioning (Skinner) - reinforced learning of new behaviors, not just shaping reflexes
- different reinforcement intervals have different effect
- complex behaviors are learned through more simple ones
Stimulus sampling theory (Estes) - a statistical learning theory; set of formulas and axioms
- S-R association is learned in a single trial; learning results in accumulated S-R associations
- reinforcement has to do with the performance, not with learning
- later included memory as a factor in his theory
1)
Approximate decade in which the theory was introduced
2)
Connectionism is not considered a learning paradigm, but is mentioned due to its influence on behaviorist ideas
3)
Stimulus-Response