6 Factors which predict experiences of behaviour
6.1 This chapter explores the extent to which different factors predict teacher and support staff experiences of negative pupil behaviour. The scope of the research only extended to exploring some specific school variables which might be linked to experiences of pupil behaviour. It was not intended to provide a full explanation of pupil behaviour as there are clearly many other variables that could contribute to this ( e.g. aspects of pupils' home lives or peer pressure).
6.2 By far the strongest predictor, for teachers and support staff in both sectors, was perceptions of school ethos: those who gave a poorer rating when asked to rate 'the overall ethos of your school' reported that they experienced negative behaviours more often.
6.3 Among secondary teachers, after perceptions of ethos, the next best predictors were working in a school with a higher proportion of pupils from the most deprived areas and being a less experienced teacher.
6.4 In order to identify what predicts staff experiences of different types of behaviours, we used regression. This is a statistical technique which analyses a number of different variables, controls for the fact that some of them might be linked with each other (for example, deprivation might be linked to the proportion of pupils with additional support needs), and separates out the effect of each to identify which has the biggest impact.
6.5 The following variables  were analysed:
- perceptions of school ethos
- school size
- school capacity
- school building condition
- proportion of pupils living in the 20% most deprived areas in Scotland
- proportion of pupils with additional support needs ( ASN)
- ratio of support staff to school roll
- ratio of support staff to pupils with ASN
- length of service of teaching staff
- the proportion of pupils that have additional support needs
- stage taught (primary teachers only)
- whether school in top 25% of schools for attainment (secondary only).
6.6 By far the strongest predictor, for teachers and support staff in both sectors, was perceptions of school ethos: those who gave a poorer rating when asked to rate 'the overall ethos of your school' reported that they experienced negative behaviours more often. This was the case for almost all the behaviour groupings ( e.g. low-level behaviour in the classroom, aggression and violence towards staff etc.) for each category of staff.
6.7 It is worth noting here that the effect may not be in one direction: ethos may have an impact on behaviour, and behaviour may have an impact on ethos.
6.8 Among secondary teachers, after perceptions of ethos, the next best predictors were working in a school with a higher proportion of pupils from the most deprived areas (which predicted five of the seven behaviour groupings) and being a newer teacher (which predicted three).
6.9 Among primary teachers, and both primary and secondary support staff, none of the other variables predicted more than two of the behaviour groupings.
6.10 See Annex B for Tables which show full details of which variables predicted staff experiences of the different types of behaviour, and for more details of how the analysis was conducted.
Analysis of subjects taught
6.11 We also undertook some anlaysis to explore whether secondary teachers' experiences of pupil behaviour were linked to the subjects that they taught. However, there was no clear pattern by subject.