Transfer of Training: Functional & Mechanistic Considerations
Ralph R. Miller, Distinguished Professor of Psychology,
Monday, May 8, 2017
About the seminar
Animals address ecological challenges by modifying their behavior across and within generations. The principle means by which such adjustments are implemented across generations is traditional natural and sexual selection, whereas within a generation it is learning. Critically, learning is beneficial only when transfer of knowledge from the training situation to the test situation occurs (and the contingencies that prevailed during training also apply to the test situation). Between- and within-generation tuning of animals' predisposition to exhibit transfer has resulted in roughly optimal amounts of transfer in many situations, but surely there are many instances of over- and under-generalization. I will describe factors that influence whether or not transfer will be observed. Some of these factors are readily subject to being manipulated to encourage or discourage transfer.
About the speaker
Dr. Miller's initial training through his masters was in particle physics at MIT and Rutgers. He then completed a second masters in social psychology where he focused on game theory, before switching to cognitive and behavioral neuroscience in which he earned his PhD. His early work in this last area was concerned with the neurobiological nature of memory where, like many researchers at the time, he tried to prevent the formation of memories in rats using electrical and chemical disruption of neural systems.
Dr. Miller’s central contribution was demonstrating that the resultant absences of evidence of training were due, not to failures in storing memories, but to performance failures, a view that is resurgent today. Additionally, long ago he discovered the phenomenon that today is called 'disrupted reconsolidation,' but here too, he views the loss of evidence of memory as a performance failure rather than a failure of reconsolidation. More recently, his research with rats, mice, frogs, and humans has examined the sources of forgetting and the conditions under which acquired information will be expressed. This includes the importance of the similarity of retrieval cues present at test relative to cues that were present during training. This similarity is at the heart of the question of transfer or training, which is the subject of his lecture.
EvoS 451 Spring Seminar
evos @ binghamton.edu