Information Technology Reference
In-Depth Information
Progressive increase in the learner's understanding and recall of the program is similar
to the size of a
learn unit
approach suggested by Greer and McDonough (1999). Moreover,
Swezey and Llaneras (1997) emphasized the importance of structured practice, which is a
feature of the present tutoring system, as a critical factor in models of training. The full
instruction set presented to the learner is available for observation within the online tutor,
which is freely available for use.
3
Access to the tutoring system will show the source of data
presented here, although the interfaces accessible on the Web are subject to updates, as they
were for Study 2 in this chapter. The interfaces and operational details for Study 1, however,
are fully explained in Emurian et al. (2000).
STUDY 1: INDIVIDUAL PERFORMANCE
The first study emphasized the detailed characterization of knowledge acquisition,
testing, and retention in three students.
4
These students used the Web-based programmed
instruction tutor to learn to write a Java Applet, which is a computer program than can run
in a browser on the World Wide Web. The design of the tutoring system is a synthesis of
principles of programmed instruction (Holland, 1960; Skinner, 1958), verbal behavior (Skin-
ner, 1957), verbal learning and memory (Li & Lewandowsky, 1995), the elaboration theory of
instruction (Mayer, 1981; Reigeluth & Darwazeh, 1982), practice and retention (Durham &
Emurian, 1998), and instructional design (Tennyson & Elmore, 1997). The system bears
similarities to those categorized by Federico (1999) as microtreatment approaches to adaptive
instructional systems. These principles will be illustrated in the examples presented for Study
1 and Study 2.
Learning and Retention
Task completion and mastery were accomplished by the learner's correct input entered
into the text editor emulation interface, which was the final interface in the tutor and the one
into which the learner wrote the entire Java program. The performance measurements were
taken on five separate occasions, where each of the first four occasions was spaced from five
to seven days apart. During Sessions 2 through 5, the learner began each session with the
item learning interface. At the conclusion of the fourth session, the learner ran the Applet
as discussed below. The Applet was not run until the fourth session so that retention and
relearning could be studied under standardized conditions involving only symbol manipu-
lations. The running of the Applet involved discussions with the expert, and these discus-
sions introduced factors that may have differed across the learners and that may be difficult
to quantify. As a test of long-term retention, the learner repeated the tutor on a fifth occasion,
which occurred at least 3 months after the fourth session. This procedure studied the behavior
of three learners with a research version of the tutor, which time stamped each interactive
event with the system into a logfile.
We studied the performance of three graduate students who volunteered for participa-
tion. These students were paid $10 for the first session and $5 for all other sessions. No
student reported previous experience with Java. Two students reported
c
and
c++
experi-
ence, and one student reported
COBOL
experience. The age and sex of each student were
as follows: S-1 (24, male); S-2 (26, female); S-3 (41, female). The third student, S-3, reported
the
COBOL
experience.
