| Year Funded: 2010 | Budget: $72,650.00 | Funding Agency: Workplace Safety and Insurance Board of Ontario |
|---|---|---|
| Title: | Firm selection algorithms – comparisons over time | |
| Category: | Intervention Research | |
| Subcategory: | Intervention Research | |
| Keywords: | Prevention Initiatives, Firm level, Targeting | |
| Link to research website: | www.iwh.on.ca | |
Issue:
A key element for managing the performance of occupational health and safety (OHS) is the ability to engage in active surveillance of how businesses are performing and being able to identify those firms where interventions are necessary. Many jurisdictions use a risk-based approach when deciding how to distribute their OHS resources. This study will investigate the impact of the method used to identify risky firms on which firms are selected and what is observed in those firms over time.
Objectives:
The overall goal of the study is to examine the nature and consequences of different mechanisms which have been used to identify firms for targeted interventions. Understanding the most appropriate firm targeting mechanisms better is critical for advancing the system’s ability to build an evidence base on intervention effectiveness. To accomplish this goal, the study builds on on-going research examining the Ontario High Risk firm Initiative (HRFI) from 2004-2008. The study will (1) identify whether the different firm selection algorithms that have been used in Ontario lead to identification of different firms or similar firms; (2) quantify regression to the mean and determine whether its impact varies across algorithms, across sectors or across firms of different sizes; and (3) determine whether regression to the mean accounts for some of the reduction in injury rates consequent to an intervention.
Anticipated Results:
To provide key information for development of best practices for firm selection or targeting using lagging indicators data i.e. those which are most consistent with subsequent performance without intervention and which reduce the likelihood of regression to the mean over the widest range of industrial sectors and workplace size.
Investigators:
Sheilah Hogg-Johnson, Benjamin C. Amick, Donald C. Cole, Cameron A. Mustard, Lynda S. Robson, Peter M. Smith, Emile Tompa, Dwayne Van Eerd (Institute for Work & Health)
