Information Technology Reference
In-Depth Information
Table 4.7 Box's test of
equality of covariance
matrices
Box's M
6.953
F
1.126
6
df1
df2
259,299.692
Sig.
0.344
of equality of error variances”. If the “Sig.” value is higher than 0.05, then the
variability in experiment's groups is about the same. In the particular experiment,
that value is 0.714 for 'grade_on_lesson1' and 0.854 for 'grade_on_lesson2'
(Table 4.8 ). Consequently, both 'grade_on_lesson1' and 'grade_on_lesson2' have
homogeneity of variances.
Then, MANOVA conducts the 'Wilks' Lambda test' to determine whether the
differences of the means are statistically significant. Lambda varies between 0 and
1. If the “Sig.” value of the test is less than 0.0005, then it means that the inde-
pendent variable has a significant effect on the dependent variable. The lambda
value of the independent variable 'group' is 0.760 and the corresponding “Sig.”
value is 0.000 (Table 4.9 ). Therefore, learners' performance on lessons 1 and 2
were significantly dependent on the use of the programming tutoring system.
Then, tests of “Between-Subjects Effects” determine how the dependent variables
differ for the values of the independent variable and if the independent variable has
a significant effect on both or only on one dependent variable. From the results
that are depicted in Table 4.10 , is concluded that the group (so, the use of the
programming tutoring system) has a significant effect on both 'grade_on_les-
son1' (Sig. = 0.000 < 0.0005) and 'grade_on_lesson2' (Sig. = 0.000 < 0.0005).
Table 4.8 Levene's test of equality of error variances
F
df1
df2
Sig.
Grade_on_lesson1
0.338
2
102
0.714
0.158
2
102
0.854
Grade_on_lesson2
Table 4.9 Multivariate tests
Effect
Value
F
Hypothesis df
Error df
Sig.
3,891.756 a
Intercept
Pillai's trace
0.987
2.000
101.000
0.000
3,891.756 a
0.013
2.000
101.000
0.000
Wilks' lambda
3,891.756 a
Hotelling's trace
77.064
2.000
101.000
0.000
3,891.756 a
77.064
2.000
101.000
0.000
Roy's largest root
Pillai's trace
0.241
6.978
4.000
204.000
0.000
Group
7.440 a
0.760
4.000
202.000
0.000
Wilks' lambda
0.316
7.896
4.000
200.000
0.000
Hotelling's trace
16.026 b
Roy's largest root
0.314
2.000
102.000
0.000
a Exact Statistic
b The statistics is an upper bound on F that yields a lower bound on the significance level
 
Search WWH ::




Custom Search