Modeling Individual Differences in Programming Skill Acquisition
Abstract
This paper investigates individual differences in learning Pascal programming skills using an intelligent tutoring system for learning criterion task. Data were collected from 260 subjects in the following categories: (a) incoming general and domain-specific knowledge; (b) cognitive ability measures (working memory capacity and information processing speed); (c) learning process measures (declarative and procedural); (d) transfer measures (retention, application, and generalization). Casual models of learning were tested linking incoming knowledge and abilities to learn processes and transfer. Finding showed that a large proportion (86%) of the learning criterion variance(transfer) was accounted for by just three factors: Working-memory capacity, domain-specific knowledge, and declarative-learning efficiency. Certain learning behaviors were also investigated in relation to transfer. Some influenced transfer performance early in the tutor while others had more impact later on. Implications of these finding are discussed in relation to individual differences research as well as to intelligent tutoring system design issues. (kr)
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1990
- Accession Number
- ADA229816
Entities
People
- Patrick C. Kyllonen
- Valerie J. Shute
Organizations
- Air Force Research Laboratory