Transfer of Learning from a Modern Multidisciplinary Perspective

• #public
• published 2005
• by Jose P. Mestre
• reading log
  • April 15th, 2020 until Historical Theories
  • April 16th, 2020 Trying to do 20 minutes, arrived at Chapter 5, p xxi
• Introduction
  • Transfer is a term that describes a situation where information learned at one point in time influences performance on information encountered at a later point in time.
  • Harald Høffding (1892)
    • learning event A
    • stored internal representation when learned a
    • internal representation of response to learned event b
    • observable response B
    • chain A-a-b-B
      • event A triggers a leads to B
    • what if you encounter A'? Will it still trigger a? How likely?
      • ​Isn't it also about how b can be applied in a different way to a slightly different problem? You wouldn't want an identical response? #q
    • Ideal
      • one learned event facilitates further learning
      • extend the range of events to which learned material can be applied
  • Types of transfer
    • Specific transfer, Near transfer and Literal Transfer are similar
    • [[Lateral Transfer and Vertical Transfer distinction]] made by Robert M. Gagné
      • Lateral Transfer
        • skill or knowledge influences more complex skill or knowledge later - precursor
        • instruction should be arranged to take advantage of hierarchical structure of knowledge to be learnt
          • part of movement of
          • task analysis
            • begin with [[terminal learning goals]]
            • move backwards until you encountered entering behaviours
              • skills students had acquired before entering school
          • mastery learning
      • Vertical Transfer
        • "generalization that spreads over a broad set of situations at roughly the same level of complexity"n
        • not well defined
    • [[Specific transfer and Nonspecific transfer distinction]]
      • Specific transfer
        • clear similarity between stimulus complex in situation A and B
      • Nonspecific transfer
        • no obvious relationship, but acquisition of one influences the acquisition of the other
          • for example learning to learn (monkey experiments)
          • poorly defined
    • [[Near transfer and Far transfer distinction]]
      • Introduced by Richard E. Mayer
        • level of similarity between conditions of original learning and in transfer learning (​how is this different from Specific transfer?)
      • James M. Royer
        • school-learned events and out-of-school events - applying something learnt at school in an out-of-school context is far transfer
      • very difficult to define
      • early thought: the far extreme of stimulus generalization gradient that will still activate a learned response
      • category of real-world problems that could be solved through use of a mathematical operation like multiplication
    • [[Literal Transfer and Figural Transfer distinction]] by James M. Royer
      • Literal Transfer
        • application of an intact bit of knowledge to a new learning event
          • example: calculating the area of a rectangle can be applied to determining size of a rug for a room
      • Figural Transfer
        • involving application of a segment of world knowledge as a tool for thinking about or learning about a new problem
          • Metaphors like "Encyclopaedias are gold mines"
        • definition by example, and not very clear
    • Actor-oriented transfer
      • 
• First chapter by Daniel Schwartz, John Bransford and David Sears
  • classic views of transfer measures transfer in ways that make people look dumb
    • measure what is "transferred out" (whether people directly apply what they learned in one context to another), and not what is "transferred in"
    • follow "sequestered problem-solving" paradigm, participants tested isolated from "contaminating" influences of information (colleagues, text, electronic resources)
  • propose expanded definition of transfer: preparation for future learning (things you learnt help you learn new things more effectively) ​pro-pedeutic?
    • transfer is not as rare as previously assumed
  • two-dimensional space of innovation and efficiency
• Second chapter by Christopher Wolfe, Valerie Reyna, Charles Brainerd
  • implications of fuzzy trace theory for transfer
    • dual-process [[theory of memory]]
      • explain surprising findings about relation between memory and cognitive processes
      • two independent memory systems
        • formed in parallel, but encoded and stored separately
        • gist
          • used to reason and problem solve
          • nebulous patterns
          • underlying meanings and patterns
          • more important for transfer
        • verbatim
          • used to answer questions about literal representations of knowledgesur
          • face details
• Third chapter by David Hammer, Andrew Elby, Rachel Scherr and Edward Redish
  • transfer as a concept is not useful.
    • traditionally describes knowledge as unitary entity - learn in one context and apply in another
  • propose a manifold ontology
    • fine-grained resources that may or may not be activated by an individual in response to a context
    • learning is a cognitive state in which several resources in the student's memory are activated and applied to deal with context under consideration
      • not as acquisition or formation of a cognitive object
• Fourth chapter by Andrea diSessa and Joseph Wagner
  • view of knowledge as a complex system containing many kinds of elements and structures
    • perspective of coordination classes
      • a model of concepts as complex systems, and the principal function of the concept/system is to allow people read specific kinds of information out of situations in the world
      • possessing a scientific concept and applying it flexibly means
        • ability to coordinate large numbers of knowledge elements in various ways that accommodate wide range of contexts over which concept is applicable
  • measure of transfer
    • ability to apply knowledge flexibly across varying contexts
  • Types of transfer
    • Expert-like transfer
      • ability to compile appropriate knowledge reliably for application in specific contexts (concept projections)
    • Novices concept projections are extremely sensitive to context, and often result in unsuccessful performance.
  • Necessary research
    • Investigations that assess not just performance, but type of knowledge used by novices in early work in domain they are learning.