Citation

Cadieux J, Roy M, Desmarais L. J. Saf. Res. 2006; 37(4): 413-419.

Affiliation

Universite de Sherbrooke, Chaire detude en organisation du travail 2500, boulevard de l'Universite, Sherbrooke, Quebec, Canada J1K 2R1.

Copyright

(Copyright © 2006, U.S. National Safety Council, Publisher Elsevier Publishing)

DOI

10.1016/j.jsr.2006.04.008

PMID

17045297

Abstract

INTRODUCTION: This article describes the validation of an instrument designed to conduct an Occupational Health and Safety (OHS) self-diagnosis using workers observations of tangible facts and actions in the workplace. The instrument puts the emphasis on observable factors that make it possible to act proactively before accidents actually occur. METHOD: Three companies were recruited from the printing sector in Quebec based on their OHS performance (low, medium, and high), their interest in the project, and their availability. The master sample was composed of 269 people. Partial least squares (PLS) and vanishing tetrad analyses were used to study the behavior of the formative scales developed. RESULTS: The results indicate that partial revision of the instrument will be necessary to reach a fully satisfying level of validity. With regards to the statistical approach adopted, the use of a macro called CTA-SAS 2.0 developed by Bollen and Ting [Bollen, K.A., & Ting, K.-F. (1993). Confirmatory tetrad analysis. In P.V. Marsden (Ed.), Social methodology.Vol.23 (pp.147-176). Cambridge, MA: Blackwell, 23] has proven to be effective in providing a suitable statistical treatment of the latent formative variables on which our instrument's development is based. CONCLUSION: The results suggest that the instrument appears to be relatively robust and appropriate to diagnose OHS leading measures. IMPACT ON INDUSTRY: In the long run, this instrument will make it possible for companies to draw a more refined picture of OHS evolution, even in the absence of occupational injuries.

Language: en