School climate has been of interest internationally for a number of years not only because school climate has been linked to the effectiveness of the school but also to learner achievement. School climate was the focus of this research study and in particular the development of a school climate questionnaire for learners in Pretoria schools. Prominent factors influencing school climate were identified from literature and six of the more prominent factors were selected for study namely cohesiveness, trust, respect, control, violence and physical infrastructure. These were conceptualised in terms of systems theory using an input-throughput-output model. Input into the system is the learners, educators, principals, policies on school-level as well as policies on National level and resources. The throughput was considered as the process of interaction between the learners, educators and principals and how this influences cohesiveness, trust, respect, control, violence and physical infrastructure. The interplay results in behaviours, perceptions and attitudes of the principal, educators and learners, which influences the atmosphere within the school (output). Survey research was undertaken in order to collect data on the six factors from the perspective of the learner. A self-administered questionnaire was used and was developed based on numerous school climate instruments.
These instruments were studied and items associated to the factors were identified for possible inclusion in the questionnaire. The items chosen were then taken and rephrased to make them relevant for the South African context. Experts in the fields of psychology, education and instrument development rated the items in terms of appropriateness, relevance, language and readability. The comments were then included and the questionnaire piloted in one school in Pretoria. Based on an initial analysis minor changes were made to the questionnaire, which was then administered in three schools in and around Pretoria. In total 608 learners participated in the study, 166 learners in the pilot study and 442 learners in the main study.
Some problems were identified with the questionnaire, which included the language of the items and that of the learners. The instrument was found to have face and content validity. The initial reliability analysis indicated that some of the factors attained reliability coefficients that were lower than the set limits. As a result item-total analysis was undertaken and it was found that certain items did not correlate well with the scale. A factor analysis was also undertaken for further scale development. Five factors were extracted using principal components analysis; the previously conceptualised factors were incorporated in different ways than anticipated. These factors were conceptualised and named Interaction, Cohesion, Learning environment (which is on a classroom level) and Resources. The only factor that concurred with the original conceptualized factors was Violence. The developed questionnaire clearly depicted these individual aspects of school climate and could distinguish between the different school contexts.