Information Technology Reference
In-Depth Information
most pronounced between the Pre-Tutor and Post-Tutor occasions, and only a few learners
thereafter were not able to produce the serial stream accurately. This improvement in
performance was observed in relationship to the enhancements to the tutoring system, which
included three presentations of the brief row tutor interface. It should also be noted that the
tutoring system was used only for the first technical exercise, and compile-time and run-time
error interpretation was taught throughout the remainder of the course. Although the writing
of the code in the transfer assessment constituted rote recollection (“regurgitation”), the
student's performance on that task was mindful, informed, purposeful, and skilled recollec-
tion.
Rather than concluding that the transfer performance failures are always problematic,
it may be more reasonable to suggest that even a professional programmer's “baseline” of
code construction contains errors that might be readily overcome with the feedback provided
by a system during program development. In contrast to such an approach, this study
focused on the serial stream of code as it might be mastered in a textual learning paradigm
(Li & Lewandowsky, 1995), and it would be informative to compare the present outcome with
a baseline of errors exhibited by expert programmers.
CONCLUSION
This chapter presented individual and classroom performance on a Web-based tutoring
system for acquiring knowledge of a simple Java Applet. The instructional design of the tutor
followed programmed instruction principles, which were supplemented with students'
personal interactions with an expert and collaborations with peers. In contrast to passive
online tutorials that display information,
5
no matter how skillfully organized and delivered
within a hypermedia environment, the present tutoring system required learners to actively
construct correct responses during the learning experience. Furthermore, the size of the
learned and constructed response was systematically increased, until the final response was
the production of the entire Java program. The programmed instruction component was
augmented by a discussion with an expert that culminated in the learner's running of the
Applet on the World Wide Web. These experiences together characterize many of the
features of a Personalized System of Instruction (Keller, 1968).
The outcomes for the three individual learners together with the classroom observa-
tions indicate the importance of repetition in acquiring competency in at least the domain of
writing an Applet from memory. Learners in the individual and classroom observations
completed the full version of the tutor and the brief row tutor or both on four separate
occasions that were similarly spaced apart over a four- to five-week interval. All three
individual learners showed performance improvement across the first four learning occa-
sions, and for the class of students, nine of the 11 Completers and three of the six
Noncompleters wrote the Applet correctly in the final assessment questionnaire that
followed one use of the full tutor and three completions of the brief row tutor. Although the
dependability and generality of these observations require intersubject and systematic
replication (Sidman, 1960), this tactic to gain confidence in the effectiveness of a tutoring
system and its enhancements offers an alternative to null hypothesis testing with “control
groups” by relying on fine-grain performance records and interbehavioral processes ob-
served over replications.
