Information Technology Reference
In-Depth Information
Figure 14. Box-plots of confidence ratings for Completers (C) and Noncompleters (N-C)
across the four observation occasions
confidence and by shifts in the interquartile range across all occasions. A test of trend,
undertaken by comparing the six successive differences between medians with a population
of zeros, was significant, F(1, 10) = 8.57,
p
< 0.02. The figure also suggests that the Completers
showed greater confidence ratings, in comparison to the Noncompleters, over the first three
occasions. The difference between the medians on the Pre-Tutor assessment was significant,
Kruskal-Wallis chi-square = 4.57,
p
< 0.04. The difference between the medians on the final
assessment was not significant, Kruskal-Wallis chi-square = 0.22,
p
> 0.50. These delimited
tests were undertaken, because the sample size is small and because parametric tests may not
be appropriate for ordinal data in a factorial split-plot design.
Table 2 presents median ratings for Completers and Noncompleters for previous Java
experience, overall impression of the tutor, effectiveness of the tutor in learning Java, and
usability of the tutor interfaces. The outcomes were equivalent between the two groups, with
the single exception of the range for the effectiveness in learning Java ratings. These ratings
show that the students reported minimal Java experience prior to this course. They also show
the favorable ratings of the tutor on all other scales.
The tutoring system provided records of errors made on the item familiarity, row, and
program interfaces. Figure 15 presents box-plots of errors on the item familiarity interface for
Table 2. Self-reported ratings; median (interquartile range) [range]
Scale
Group
Completers
Non-Completers
Experience
1 (1-2)[1-2]
1 (1-2)[1-2]
Overall
5 (4-5)[3-5]
5 (4-5)[3-5]
Learning Java
5 (4-5)[3-5]
5 (4-5)[4-5]
Usability
5 (4-5)[4-5]
5 (4-5)[4-5]
