Information Technology Reference
In-Depth Information
Discussion
Three individual learners constructed a Java Applet from recall on five separate
occasions, as the final performance requirement in a Web-based tutoring system. All learners
showed performance improvement over the first four sessions. The most pronounced
improvement was noticed between Session 1 and Session 2, a general outcome that is
consistent with practice effects and the power function of learning (Lane, 1987; Wickens &
Hollands, 2000, Chapter 7). Furthermore, each learner showed an idiosyncratic acquisition
process that culminated in mastery of the Java code. It is the individual acquisition process
that is often obscured by relying on group averages to characterize performance. However,
the common features of learning, in terms of improvement over successive sessions, were
similar to acquisition curves found in many different knowledge domains (Swezey & Llaneras,
1997). And the fact that all learners showed at least some forgetting between successive
session occurrences, despite the accurate writing of the Java code at the conclusion of each
session, indicated the importance of repetition in achieving a steady-state asymptote in
performance on the task. Shute and Gawlick (1995) also demonstrated relearning of a task,
after a 2-year delay interval, as an index of initial learning effectiveness. Additionally, the fine-
grain data records over successive sessions may lend themselves to interpretation in terms
of a power integration diffusion model of performance following breaks in learning (Sikstrom
& Jaber, 2002). Finally, the pronounced drop in performance effectiveness that was observed
during the fifth session suggested that even a well-practiced skill may suffer degradation if
it is not regularly used.
STUDY 2: CLASSROOM PERFORMANCE
The validation of the tutoring system, by studying fine-grain performance with a small
sample of learners, provided the occasion for adopting this approach in the classroom. Our
subsequent research reported the classroom application of this programmed instruction
tutoring system (Emurian & Durham, 2001). The latter research showed that experience with
the tutoring system, together with subsequent classroom instruction, produced dependable
improvements in students' self-reports of confidence in the use of Java symbols. It was the
case, however, that the students' writing of an error-free Java Applet did not always carry
over from the tutor to a later assessment occasion. In fact, only two of 12 learners observed
in that study were able to write the program correctly immediately after completing the tutor.
This outcome was observed despite the fact that one error-free production of the program
was required to exit the tutor. Although the tutor presented explanations of the code, together
with revised multiple-choice tests of the meaning of individual items of code and rows of code,
a more robust transfer of training was anticipated between the tutor experience and
subsequent transfer assessments of retention of the program.
Against that background, Study 2 within this chapter intended to show enhancements
of student learning and retention of the Java Applet under consideration by using the
previous classroom study results as a baseline for comparison. This approach to program-
matic improvements follows the experimental design methodology of systematic replication
(Sidman, 1960). In this research methodology, enhancements are undertaken to select
independent variables under analysis with the objective to potentiate an effect that has
practical rather than statistical significance. In the present circumstance, it is the objective
of the enhancements to improve all students' constructions of an error-free Java program
