Safety Climate Measurement: an Empirical Test of Context-Specific vs. General Assessments

Abstract

Safety climate researchers develop and use both general and industry-specific safety climate measures. Theories about language comprehension suggest that context facilitates meaning; however, the relative value of context-specific safety climate measures in the prediction of safety outcomes is an empirical question that has not been rigorously tested. The purpose of the present study was to provide a rigorous comparison of context-specific vs. general safety climate measures. Seven hundred forty-six university laboratory personnel from five different kinds of research labs (i.e., animal biological, biological, chemical, human subjects/computer, or mechanical/electrical) completed contextualized safety climate measures, a general safety climate measure, and measures of other safety-related constructs. Measurement equivalence analyses indicated that the general safety climate measure was not equivalent across the five lab types. Hypothesis testing revealed that contextualized information was most helpful when included in safety climate measures for less, rather than more, safety-salient contexts, but overall, there was relatively little difference in the validities for general and context-specific measures. Results suggest that context has a small influence on how individuals respond to safety climate measures and provide guidance for researchers/practitioners when deciding between using industry-specific or general safety climate measures. It appears most beneficial to use industry-specific measures when examining safety climate in a less-safety-salient context. This study offers one of the first empirical tests of a contextualized safety climate measure involving a rigorous, unconfounded comparison of five context-specific safety climate measures with a general measure.