Internet Learning Volume 6, Number 2, Fall 2017/Winter 2018 | Page 34
The Effect of Term Length on Student Achievement in Online College Algebra
confirm that the sample data do not follow
a normal distribution.
Since the data do not meet the
assumption of normality for the parametric
independent samples t-test,
Corder and Foreman (2014) recommend
turning to a non-parametric test.
The non-parametric equivalent is the
Mann–Whitney U-test. With this test,
the samples are combined and rank-ordered
together to see if the values are
randomly mixed in the rank ordering
or if they are clustered at opposite ends
(Corder & Foreman, 2014). The null
hypothesis is that there is no tendency
of the ranks of one method to be systematically
lower or higher than the
other (Corder & Foreman, 2014). Table
3 shows the SPSS output for the Mann–
Whitney U-test.
Table 3: Mann–Whitney U-test
Ranks
Score
Term N Mean rank Sum of ranks
0 691 411.69 284,481.00
1 121 376.83 45,597.00
Total 812
Test Statistics a
Score
Mann–Whitney U 38,216.000
Wilcoxon W 45,597.000
Z
−1.512
Asymp. Sig.
(two-tailed) 0.131
Because the sig. value (0.131)
is greater than the α value of 0.05, the
null hypothesis fails to be rejected. This
indicates that neither the 8-week term
nor the 16-week term yields higher final
exam scores.
To see if the inclusion of the
outliers influenced the results, another
Mann–Whitney U-test was performed
with no outliers present. The results are
presented in Table 4.
33