Date of Award


Degree Type


Degree Name

Doctor of Philosophy


Faculty of Education

First Advisor

Professor Sybe Jongeling

Second Advisor

Dr Amanda Blackmore


This study investigated the presence of occupational stress among teachers. It did not set out to identify and explain variables associated with stress among teachers, rather the study focussed on the multitude of variables identified in the literature and sought to include these in a more extensive causal model. The study was conducted in two main stages. Firstly, a cross-sectional survey investigated the presence of stress among 230 teachers, as measured by Psychological Stress, Physical Health, Job Satisfaction and & desire to Leave their Job. The survey obtained information on stress outcome variables (Psychological Stress, Physical Health, Job Satisfaction and Wanting to Leave), biographical information, personality (Hardiness, Type A Behaviour, Locus of Control, Extraversion and Neuroticism), psycho-social variables, (Social Support, Problem Solving, Emotional Coping mechanisms and Self Esteem) and work and life stressors (Role Conflict, Role Ambiguity, Job Responsibility, Job Future Ambiguity, Underutilization of Skill, Inequity of Pay, participation in Decision Making, Administrative Support, Relationships with Peers, Extra Work, Wanted Extra Work, Workload, Work Hours and major Life Events). Causal models using path analysis were then generated to account for the relationships found within the data. Secondly a longitudinal study over six months was conducted on 242 teachers. The causal models generated in the first study were re-tested on this second group of teachers, both cross-sectionally and longitudinally. Furthermore the importance of existing levels of stress in the prediction of future stress was assessed. Finally a three year follow up was conducted on the survey's participants. The results of the investigations revealed that:-(i) the utility of demographic information in the stress process was inconsistent. (ii) that among measures of work stressors there are replication and redundancies. (iii) that the different stress outcome measures had different predictors. (iv) that the best fitting causal models for the stress process were direct effect models. (v) that existing stress levels am an important factor in the prediction of future stress levels. (vi) that those teachers who indicated a desire to leave teaching and/or were experiencing psychological stress, were more likely to leave teaching three years later. (vii) that there is a need to standardize both the outcome and predictor variables used in stress research. The implications of these results were then discussed as were areas for future research.