Abstract
The present study consisted of a multivariate analyses of approximately 3500 accident cases taken from a large petro-chemical complex outside of Houston, Texas. The accident reports were coded using a system similar to that recommenced by the National Safety Council, and from it frequencies and rates were generated for analyses. Log-linear analysis was conducted to test for age, job and task effects on accident rates as well as interactions between them. All models were found to be significant at P.05 level, indicating that age, job and task all have important impact in determining accident rates. The significance of the interactions show that age has a differential impact on the rate of accidents found within given jobs and tasks.
Get full access to this article
View all access options for this article.