Abstract
Although there has been considerable interest in safety leadership, relatively little attention
has been given to the factors that determine safety leadership in the construction industry.
The effective functioning of the construction industry is central if the South African
government is address the challenges related to the provision of infrastructure and achieve its
development goals. The construction industry is characterised by a wide range of diverse
activities that take place in a highly complex and hazardous environment. The industry is also
known for its high levels of injuries and fatalities. This study focused on the variables that
influence the leadership competencies required to strategically improve safety performance in
the construction industry. The study also aimed to construct and evaluate a measuring
instrument or tool that could be used to develop a competency framework for safety
leadership in the construction industry. It is essential that the construction industry take
congnisance of the rapidly changing environment in which it operates and, in particular, the
current and evolving economic, legal and regulatory changes that affect it. It was against this
backdrop that the literature review was conducted and the existing body of relevant
knowledge explored.
In order to realise the empirical objectives of the study, a pragmatic research paradigm, using
a methodological triangulation approach, was adopted. This approach involved the use of
both a qualitative research method, namely, interviews, and a quantitative research method,
namely, a questionnaire. This approach was deemed appropriate as the pragmatic paradigm
applies all research which involves understanding a specific research problem. Combining
quantitative and qualitative data is advantageous in that it enables the researcher to develop a
more complete understanding of the research problem than would otherwise have been the
case, to compare and validate the research results, to provide illustrations of context for the
trends discovered and to examine processes or experiences together with outcomes. The
safety surveys conducted were characterised by acceptable reliability, strong face validity and
some predictive utility. Predictive utility was supported in that the regression models
obtained adjusted r-square values of up to 0.47 , indicating that as much as 47% of the
variation around the mean of the dependant variables could be explained by the model.
Correlational analyses were conducted to determine whether a positive relationship exists
between the various constructs and to assess the scales which accounted for the most variance
in employee perceptions of safety culture and management commitment...
D.Phil. (Mechanical Engineering)