Information Technology Reference
In-Depth Information
passes, and that may have driven down the number of times that the learner would use the
selection to observe the code.
Three of the 11 Completers selected one review on the program interface, and they made
no errors. Two other Completers made two errors each on the final program interface, and
neither of these learners selected the review option.
Figure 17 presents the total number of correct Java programs that were written into the
WebCT questionnaire over the four assessment occasions for both groups. The accuracy
of the programs was judged independently by two Java instructors who showed 100%
agreement. There was no feedback given for incorrect responses, and there was no
opportunity to compile the code to test for compilation errors and to revise the code. The
assessment consisted only of the learner's memory in entering a correct serial stream of Java
items that would produce a Label object in an Applet container.
In Figure 17, it is shown that no learner wrote a correct program during the Pre-Tutor
assessment. Immediately after completing the tutor, which required one accurate construc-
tion of the entire program, nine Completers entered the code correctly during the Post-Tutor
assessment, and one Noncompleter entered the code correctly. This shows the skill
acquisition effects afforded by the succession of tutor interfaces, even when the final
program interface was not completed by all students. After receiving classroom instruction,
which involved writing and compiling the program and running the Applet, eight Completers
and two Noncompleters entered the code correctly during the Post-Applet assessment. At
the end of the course, nine Completers and three Noncompleters entered the code correctly
during the Final assessment. There were only two instances in which a correct program written
by a learner on an early assessment was not written correctly on a later assessment.
Discussion
The present classroom study followed the methodology of systematic replication
(Sidman, 1960), and it was intended to extend and clarify the outcomes observed in the prior
Figure 17. Total number of correct Java programs that were written into the transfer
assessment questionnaire for Completers (C) and Noncompleters (N-C) across the four
observation occasions
