Data Collection

The main data collected in the experiment are the students' attendances along several courses. For each student, we collected the exact chair (identified with the row and column) where they sat in each session. Moreover, at least one official exam was registered for each course. We combined the final marks with the other available data to generate a number of reports with statistics and other calculations that are the basis of more advanced statistics shown in pages results and further statical analysis.

One of the biggest challenges was to obtain statistics not related to individual students but, on the contrary, associated with the different physical positions that can be occupied inside classrooms. To achieve this goal, we designed the experiment in a novel way: we collected the data linked to each individual chair of each classroom instead of analyzing the students. Therefore, the data linked to each chair have been produced by combining the data collected every time a student (not necessarily the same one) sat in this particular chair. All data were later combined to produce extrapolative results. The collected data are the following:

Average mark of a chair:
It represents the average mark a specific chair got in a particular exam. It was obtained by adding up the marks of all students that occupied that chair (if the same student occupied the chair several times, their mark was also counted several times) and then dividing the result by the number of times this chair was occupied (hence, we get the average mark of the individual occupation of this chair).1 We can see specific examples of this in Figure 1. Observe that each chair (not each student) is labelled with a mark. Observe also that each computer is shared by two students in the labs. The blank chairs were never occupied by any student in any session. Lecture halls and labs have a symmetric and proportional distribution of chairs, thus, row 2i is twice as far to the professor as it is row i. The reader should not extract conclusions from these figures, because they just show unprocessed data from two examples of courses. These data (together with the rest of courses) are mixed and statistically analyzed in results page.
Times a chair is used:
It counts the number of times that (possibly different) students occupied the chair during the course. Figure 2 contains examples of these counters.
Times a chair is used by students who gave up the course:
It represents the total amount of times a chair was used by a student who gave up the course (those that did not take the exam). Examples are shown in Figure 3.


PIC
(a) Course: PRG - Group: PL1
PIC
(b) Course: GUI - Group: PL5
Figure 1: Two (real) examples of average marks of chairs in a group
PIC
(a) Course: PRG Group: PL1
PIC
(b) Course: GUI - Group: PL5
Figure 2: Two (real) examples of number of times each chair has been occupied
PIC
(a) Course: PRG - Group: PL1
PIC
(b) Course: GUI - Group: PL5
Figure 3: Two (real) examples of number of uses of chairs by students who did not take the exam