Information Technology Reference
In-Depth Information
discussed. The students then compiled the Java code and ran the Applet in a Netscape
Communicator© browser by accessing the HTML file as a URL on the Web. To foster a
collaborative learning environment, the students were encouraged to help each other and to
seek help from the instructor and course assistant as needed. This part of the classroom
experience was based upon the Personalized System of Instruction (Keller, 1968), which
included interpersonal interactions as a further means of learning and competency testing
(Ferster & Perrott, 1968). After all students ran the Applet on the Web, they again completed
the confidence ratings and the writing of the Applet code in the WebCT text box. This identical
assessment was repeated during the 14th class, the final class of the course.
The enhancement to the version of the tutoring system under consideration also
included the addition of a
brief row tutor
interface. This interface was similar to the first pass
through the
row
interface in the full tutoring system. Whenever an error was made on any
row in the brief row tutor interface, the learner was given the opportunity to view the correct
code for that particular row and to enter the code repeatedly until the row code was accurate.
There was also a multiple-choice test on each row, and the test was identical to the one given
in the full tutoring system. The purpose of this brief row tutor interface was to provide
additional rehearsal and overlearning of the Java Applet at different temporal occasions
throughout the course. The brief row tutor interface was administered on classes three, seven,
and ten of the 14-class course. It was administered at the beginning of these classes, and 30
min were allotted for the learners to complete the brief row tutor. This was sufficient time for
all students to complete this interface.
In pursuing this classroom study, at least two constraints were evident. First, we wanted
to offer the tutor to students, rather than to research “subjects.” Accordingly, the tutor was
administered as a class exercise, and the procedure was exempted by the Institutional Review
Board, because it dealt with instructional technology. Second, we were bound by the length
of the class. This led to a consideration of performance differences that manifested
themselves between those students who completed all stages of the tutor and those students
who did not. Although students might have been encouraged, if not required, to complete
the tutor outside of class, experience suggested that this option would introduce uncon-
trolled factors into the interpretation of the results.
Results
At the conclusion of the class time allotted for completing the full Java tutor during the
first class, 11 students (“Completers”) had finished all parts of the tutor, and six students
(“Noncompleters”) were still working on the row (Stage 6) or program (Stage 7) interface. The
data analysis is reported as a between-group comparison between these two groups of
students. The rationale for this approach was to investigate potential early indicators that
a student would not complete the tutor in the time allotted. This evidence might prove useful
in future tutor enhancements that require different degrees of rehearsal or practice in
relationship to the knowledge and skill levels of the learners, perhaps detected early in the
tutor exercises.
Figure 14 presents box-plots of confidence ratings for both groups across the four
assessment occasions. The data are based upon the median confidence rating for each
student across the 21 items evaluated. The figure shows graphically that the ratings of
confidence, as evidenced by the median and interquartile range values, increased for both
groups across all occasions. This is evidenced graphically by the increases in median
