The authors recognize the following limitations of the study
that adversely affect the internal validity, external validity, and
reliability:
• A relatively small sample size that limits the ability to generalize
the results to the entire construction industry. A major reason
that the sample size was small was the reluctance of the contractors
to provide sensitive data. The authors recommend further
study with a larger number of projects to address this limitation.
• This data set had some missing data because some contractors
do not measure or track all values requested. Most commonly,
first-aid injuries and the number of items requiring rework were
the least-tracked data, whereas OSHA recordable injury rates
and the dollar value of rework were the most-tracked data.
• A disproportionate proportion of projects (47%) were located
in Colorado where the study was conducted. This limitation
reduces the ability to generalize the results to other states.
However, there is no reason to believe that Colorado is a special
population. Thus, it is reasonable to believe that the results can
be generalized.
• This research does not consider other safety interventions
implemented by the contractor or the quality of their safety
program.
• The size of projects varied widely, which may limit the ability to
generalize the findings to a particular portion of the construction
industry.
• The sample size used to examine some specific relationships
(i.e., Figs. 1 and 2) are very small and, thus, should be considered
suggestive only and as prompts for future inquiry.