In the experiment described below the author was interested in whether a specific mathematical method makes a significant difference on students’ performance. The experiment was carried out at “Pia Bratianu” School of Bucharest during the second semester of school year 2012-2013. The sample was composed by students of three grades 8 classes who were supposed to have an increased motivation for learning Mathematics because of their incoming national grade 8 examinations (called national evaluation) as well as the reputation of the school. The experimental design consisted of three groups: class VIII_A (group 1), class VIII_B (group 2), and class VIII_C (group 3). Group 2 represented the experimental group while groups 1 and 3 were considered the control groups. The reason we chose two control groups instead of one is twofold. One the one hand, we did not assigned randomly the students to the respective groups but we preferred to let them learn into their natural classroom environment. One the other hand, the experiment was performed continuously during one semester which in case of redesigning the classes (e.g. assigning randomly and equally the students to the three groups) the educational process of the entire school would have been disturbed. Moreover, in order to reduce more the sampling errors, all three groups were tested initially (e.g. we used a more reliable external assessment organized by the Ministry of Education, named “pre-testing”or “simulation” instead of a teacher made pre-test).
The design of the experiment is shown in table 1 where by “treatment” and “observation” we meant the new teaching method used (e.g. Math Project method) and the tests applied, respectively.
Why all of these are carried out occasionally in the classroom?