The researchers purposively selected schools with two form one streams and above from a
list of co-educational district secondary schools in Nakuru district. Four schools, E1, C1, E2
and C2 were randomly selected from the list and then randomly assigned to 2 experimental
and 2 control groups as required by quasi-experimental Solomon Four Group Design. If a
selected school assigned to an experimental group had more than two classes, all the students
were exposed to the treatment but two streams each were randomly selected for analysis.
According to Mugenda & Mugenda (1999) at least 30 subjects per group is suitable for
experimental research. The single classes used for the exercise in the four schools had
between 35 and 45 students each and the actual sample size obtained for double streams in
the four schools was 323. This consisted of 161 male and 162 female students. The
experimental groups had 85 male and 89 female students while control groups had 76 male
and 73 female students on which analysis was conducted in the study.