Comparing recall vs. recognition measures of accident under-reporting: A two-country examination

Abstract

A growing body of research suggests that national injury surveillance data significantly underestimate the true number of non-fatal occupational injuries due to employee under-reporting of workplace accidents. Given the importance of accurately measuring such under-reporting, the purpose of the current research was to examine the psychometric properties of two different techniques used to operationalize accident under-reporting, one using a free recall methodology and the other a recognition-based approach. Moreover, in order to assess the cross-cultural generalizability of these under-reporting measures, we replicated our psychometric analyses in the United States (N=440) and Italy (N=592). Across both countries, the results suggest that both measures exhibited similar patterns of relationships with known antecedents, including job insecurity, production pressure, safety compliance, and safety reporting attitudes. However, the recall measures had more severe violations of normality and were less correlated with self-report workplace injuries. Considerations, implications, and recommendations for using these different types of accident measures are discussed.

Keywords: Accident under-reporting; Measurement validation; Surveillance.