Internet Learning Volume 6, Number 2, Fall 2017/Winter 2018 | Page 31

Internet Learning Journal
Data Analysis and Findings
A
histogram and basic descriptive
statistics for the 8-week
sample are presented in Figure
1. The final exam scores in the sample
range from 12 to 100. The histogram
shows that the data are skewed to the
left with a mean score of 78.87.
The histogram and basic descriptive
statistics for the 16-week sample
are presented in Figure 2. The final
exam scores in the sample range from
20 to 100. The histogram shows that
the data are slightly skewed to the left
with a mean score of 75.81.
Preliminary inspection of the
two data sets reveals that the difference
of the means is 3.06 points, where
the mean score of the 8-week group is
slightly higher than that of the 16-week
group. The standard deviation of the
16-week group is slightly higher than
that of the 8-week group, indicating
that the scores of the 8-week group
are more closely clustered around the
mean.
Possible outliers were determined
at three standard deviations
above or below the mean. In the 8-week
group, this range was from 26.046 to
131.694. Clearly, there were no scores
above 100, and there were 12 scores
that fell below 26 points. In the 16-week
group, the ±3 standard deviation range
was from 16.92 to 134.7. No scores fell
outside of this range.
It was decided to keep the twelve
scores from the 8-week and not remove
them as outliers. This decision
was made because the scores represent
content knowledge on the exam. While
low, in terms of relation to the mean,
the scores are still meaningful and inform
the research question.
Originally, I planned to use an
independent samples t-test to compare
the mean final exam scores of the
two groups. In order to conduct an
independent samples t-test, multiple
assumptions must be met. These assumptions
include having a dependent
variable that is measured on a continuous
scale; an independent variable that
contains two categorical, independent
groups; independence of observations;
no significant outliers; for each group
of the independent variable, the dependent
variable should be approximately
normally distributed; and homogeneity
of variance must be present (Gall et al.,
2003). Each of these assumptions will
be discussed in detail below.
For this study, the independent
variable was the term length, either 8
weeks or 16 weeks. The dependent variable
was the score on the final exam.
Scores ranged from 0 to 100 (where
partial credit in varying amounts could
be awarded for each problem) and, as
such, were measured on a continuous
scale. There exists independence of observations
since there was no relation
between the participants in the two
groups nor in their test scores. Possible
outliers were discussed above. Multiple
methods were used to test for normality
of the data, all of which are described
below.
30