Journal of Academic Development and Education JADE Issue 8 | Page 20

20 | JADE
ARTICLE #2 | 21
CARLOS TURRO, IGNACIO DESPUJOL & JAIME BUSQUETS
A CASE STUDY IN LARGE SCALE VIDEO RECORDING USING OPENCAST
Results and Discussion
In order to get an evaluation of the lecture capture experience
we selected 10 courses of the first semester of the academic year
(September to January) that have been using lecture recording in all
or nearly all lectures. These 10 courses belong to different faculties
and disciplines, ranging from an introductory course in Mathematics
to different more advanced engineering topics. There are 3006
students enrolled on those courses.
The question that we try to solve on this analysis is if there is
a measurable effect on assessment grades due to the use of the
Lecture Recording tool. So we are going to divide the students in two
groups: those that have viewed at least once any video and those
that don’t. While further classification could be done, we believe that
this simple separation address more clearly the question proposed.
Then, for those 10 courses we have 2034 students that pertain to
the “viewers” group, and had accessed the platform to see at least
one video, being 67.6% of the global group. It is also worth noting
that students didn’t have any special pressure in viewing videos, it is
completely up to them.
The details of “viewers” and “non-viewers” are depicted on table I.
There are courses (5,7,8) with nearly full attendance where there are
others (6,10) with quite low usage. The overall figure is related to the
whole number of students.
Now we are going to focus on the actual marks of the students at
the end of the semester. Marks are also given on Table 1 and run from
0 to 10, being 10 the highest.
their knowledge to be helped by only having the recordings avail-
able.
Course
Viewer
Students
Non-viewer
students Percentage
Viewers’
marks
Non-
viewers’
marks
1 433 89 82,95% 5,82 5,76 1,09%
2 324 78 80,60% 6,50 6,46 0,63%
3 308 88 77,78% 6,28 6,36 -1,25%
4 298 78 79,26% 7,38 7,20 2,38%
5 264 1 99,62% 6,46 8,10 -25,39%
6 179 472 27,50% 5,78 5,37 7,15%
7 106 2 98,15% 7,13 5,50 22,83%
8 47 0 100,00% 6,62 - -
9 38 30 55,88% 7,66 7,78 -1,57%
10 37 134 21,64% 6,50 5,92 9,02%
Overall 2034 972 67,66% 6,44 5,80 9,91%
Table 1: Results of the Lecture Capture experience
On the following page we see that there is some disparity between
groups, but on the overall sample we have a 9% difference in terms
of using the Lecture Recording system, that given the size of the
sample can be considered like a positive indicator of correlation.
While this correlation is interesting, we can get a better knowledge
of what is happening by looking at the distribution of marks. On
Figure 1 we show a histogram of the marks over the full sample of
students.
Looking at that figure we see that there is a displacement of the
peak of the graph between viewers and non-viewers, while the left
and the right side keep pretty similar. Our explanation on that is that
Lecture Recording usage helps more to middle level students than
to high and low level students, probably because high level students
already know the topic enough and low level students are too far on
Difference
Figure 1: Academic results of Opencast Lecture Recording