Design for worksystem safety using employees' perception about safety.

In this paper, the predictors of work injuries based on Leamon's Man-Machine model are identified in a sociotechnical framework. Several hypotheses are developed and tested to describe the accident/injury phenomena in mining worksystems. Possible designs for improving work-system's safety are specified using scaled Mahalanobis distance (MD). A case control study design is adopted. Five variables namely, age, negative-affectivity, physical-hazards, job-dissatisfaction, and safety-practice are emerged as significant contributors to work injuries for the mines studied. Two most interesting findings obtained through this study are (i) 36% of cases (injured employees) (MD < 1) are unlucky to meet an accident and (ii) 40% of the controls (non-injured employees) (MD > 1) are lucky to be able to avoid an accident. The most probable reason for the former case is the organizational ineffectiveness while that for the latter may be risky adventures of employees which are due to lack of education, awareness, and appropriate training. Based on the MD values for cases and controls, possible design guidelines are suggested. The study categorically identifies the accident situations where engineering control, education and training, and other organizational safety measures are to be adopted.