Validation of the group nuclear safety climate questionnaire

• The main purpose of our study was to validate an instrument to measure group safety climate in nuclear power plants.
• First, we adapted it to the nuclear sector. As the original scale measured global safety, we had to adapt our measure to nuclear safety, which concentrates on protecting people and the environment from radiation risks (IAEA, 2009).
• Second, in some of the items we changed the referent from supervisor to group. Therefore, it provides a more comprehensive evaluation of group safety climate. A group safety climate scale that leaves out the perceptions of group members’ practices and behaviors related to safety offers a limited understanding of group-level safety climate.
• The Confirmatory Factor Analysis showed that the single factor was better than the two-factor and three-factor models. These results agree with those obtained by Zohar and Luria (2005), and they contrast with the three-factor model found by Johnson et al. (2007). Our measure has shown evidence of validity based on relationships with other variables (AERA, APA, NCME, 1999), since group safety climate was correlated with other variables (e.g. organizational safety climate, formalized procedures and safety behavior). Additionally, aggregation indexes were adequate, which means future studies could be developed at the group level.
• Therefore, our main contribution consists of providing a tool to measure group safety climate in high reliability organizations in the Spanish context. We have provided an instrument to measure this important factor, group-level safety climate, in high reliability organizations, where safety is considered a main issue in improving organizational excellence.

IntroductionGroup safety climate is a leading indicator of safety performance in high reliability organizations. Zohar and Luria (2005) developed a Group Safety Climate scale (ZGSC) and found it to have a single factor.MethodThe ZGSC scale was used as a basis in this study with the researchers rewording almost half of the items on this scale, changing the referents from the leader to the group, and trying to validate a two-factor scale. The sample was composed of 566 employees in 50 groups from a Spanish nuclear power plant. Item analysis, reliability, correlations, aggregation indexes and CFA were performed.ResultsResults revealed that the construct was shared by each unit, and our reworded Group Safety Climate (GSC) scale showed a one-factor structure and correlated to organizational safety climate, formalized procedures, safety behavior, and time pressure.“Impact on IndustryThis validation of the one-factor structure of the Zohar and Luria (2005) scale could strengthen and spread this scale and measure group safety climate more effectively.