RESULTS

Demographic

In total, 18 companies participated in the study. Based on the initial interview questions pertaining to general information about the company, we were able to identify the number of employees for each company. Two of the companies employed over 100 employees (134 and 150), and were therefore excluded from all further analyses. The median number of employees in the 16 remaining companies was 30, the number of employees per company ranged from 12 to a maximum of 50. The initial list of 28 companies provided by NSCSA contained 13 high claim rate companies and 15 low claim rate companies. The 16 companies included in the study consisted of ten low claim rate companies and six high claim rate companies.

Comparison between High and Low Claim Companies

Figure 3.1 presents the mean and standard deviations for the high and low claim rate companies for each of the 11 indicators included in the interview protocol. We conducted a series of Mann-Whitney U tests to examine differences between high and low claim rate companies on the 11 indicators. High and low claim rate companies are significantly different on four of the 11 indicators: incident investigation (Mann-Whitney U test; z = —2.66, p < . 05); employee involvement (Mann-Whitney U test;

Figure 3.1 Comparison between low and high claim rate

z = —2.74, p <.05); manager safety inspections (Mann-Whitney U test; z = —2.31,p < .05); and communication (Mann-Whitney Utest; z = —2.26, p < .05). We also created an average overall score for each participating company, by aggregating their score on the 11 indicators. Low claim rate companies had significantly higher aggregate scores than high claim rate companies (Mann-Whitney U test; z = — 2.66, p < .05).