How do resources (especially time available and prior knowledge) influence
learning strategies (knowledge acquisition) and later application use of
that knowledge (knowledge use)?
We could show that subjects learning with worked out examples are superior
to subjects learning by active problem solving. This effect can be traced
back to difference in the allocation of resources during learning. In the
next step we want to investigate how differences in prior knowledge and
time available influence the selection of learning strategies and problem
solving strategies.
For running appropriate experiments we use a hypertext learning system
especially designed for our purposes that deals with the mathematical field
of combinatorics. There is a learning phase using different kinds of
example problems and a test phase presenting word problems that subjects
have to solve for themselves.
We apply knowledge assessment methods developed for investigating subjects
cognitive control processes underlying strategy selection in the learning
phase and the test phase. We want to find out in which order subjects work
on different test problems and how they manage task shifts from one problem
to another. Other strategies of interest to us concern the acquisition and
use of simple or elaborated example knowledge and the acquisition and use
of abstract problem solving schemas.
Our results will be modeled using the cognitive architecture ACT-R. This
allows us to demonstrate how different aspects and regularities of human
learning behavior can be realized in computers.
|