Pask, Andrew Gordon Speedie; (1964) An investigation of learning under normal and automatically controlled conditions. Doctoral thesis (Ph.D), University of London.

Preview

Text Pask_10208501_thesis.pdf Download (62MB) | Preview

Abstract

Automatically and adaptively controlled experimental situations are defined and the use of these situations in psychology is reviewed. An abstract model is developed in torms of which it is possible to decompose an adaptive system, consisting of a subject interacting with an adaptive control mechanism, into experimentally tractable components and to decompose the skill which the subject is learning into subskills. A number of hypotheses are posed and treated experimentally in connection with a single class of perceptual motor skills. The most interesting conclusions urea (1) The simplest (single level, single subskill) adaptive system is stable, and the learning process satisfies the postulate that the rata of learning is greatest when the subject is maximally loaded. (2) It is possible to obtain stationary state measurements that characterise the learning process and that may have general predictive value. (3) Many level systems with a pair of subskills are stable (further, the higher level control is effective). (4) It is possible to distinguish and to establish systems in which the adaptive process does and does not entail subject "involvement" or "participation". (5) Certain detailed learning models of a "Cybernetic" or an “Artificial Intelligence" calibre can be tested in adaptively stabilised conditions and it is also possible to distinguish between different repertoires of problem solving procedures that satisfy the same behavioural specification.

Download activity - last month

Export as XMLJSONCSV

Download activity - last 12 months

Export as XMLCSVJSON

Downloads by country - last 12 months

Export as JSONCSVXML

Archive Staff Only

View Item