Finite Markov Learning Models for Knowledge Structures
- 176 Downloads
Some finite Markov learning models are described, which explain the transitions, over time, between the various states of a knowledge structure, from the empty state to the full domain of the structure. Any one of these models sets constraints on the probability distribution on the collection of knowledge states, thereby reducing considerably the number of parameters that need to be estimated. Such models are intended for use at an early stage in the analysis of data. The resulting knowledge structure should be regarded as tentative, and can be refined by further analysis with more demanding models.
Unable to display preview. Download preview PDF.
- Falmagne, J.-Cl. (in press). Stochastic learning paths in knowledge structures. Journal of Mathematical Psychology.Google Scholar
- Kambouri, M., Koppen, M., Villano, M., & Falmagne, J.-Cl. Knowledge assessment: Tapping human expertise by the QUERY routine. Technical Report MBS 91-20, Irvine Research Unit in Mathematical Behavioral Sciences. Irvine: University of California. (Submitted for publication.)Google Scholar
- Lord, F.M., & Novick, M.R. (1974). Statistical theories of mental tests scores. (2nd. ed.) Reading, MA: Addison Wesley.Google Scholar
- Villano, M. (1991). Computerized knowledge assessment: Building the knowledge structure and calibrating the assessment routine. Doctoral dissertation. New York: New York University.Google Scholar