In choosing the sample of interviewees we ensured that the elicited data would be representative of College One. Cohen, Manion & Morrison (2003) argue that the researcher when choosing the sample has to consider the extent to which the sample represents the whole population. When choosing the sample of College One for the study we decided to interview the college principal, all the heads of school (16 in all) and a sample of 34 teachers to form an adequate representation of the school practitioners of College One. The sample of interviewed teachers was randomly chosen from the school staff lists that the heads of school willingly supplied. We chose a representative sample of teachers on the basis of gender and teaching experience so as to ensure the right mix. The range of professional experience of interviewed teachers ranged from 2 to 40 years and their professional qualifications ranged from a teacher training college certificate to a postgraduate degree.