Yueng-hsiang Emily Huang, PhD, Oregon Health and Science University; Jin Lee, PhD, Kansas State University; Zhou Chen, PhD, University of Connecticut; Sarah E DeArmond, PhD, University of Wisconsin Oshkosh; Anna Kelly, BS, Oregon Health and Science University; Yimin He, PhD, University of Nebraska Omaha
Purpose/Objectives
The purpose of this study is to provide a Spanish-language version of Zohar and Luria’s (Zohar & Luria, 2005) commonly-used safety climate scale using a rigorous translation-back translation process.
Background
Safety climate has been defined as workers’ shared perceptions of the value and importance of safety in their organization as reflected through policies, procedures, and practices, at a particular point in time. It is well-established that safety climate is one of the best predictors of safety outcomes (e.g., frequency and severity of injuries). During the past four decades many measures of safety climate have been developed; however, one of the most popular measures has been the 32-item scale created by Zohar and Luria (2005). The scale consists of two levels, organization-level safety climate and group-level safety climate. There is significant evidence supporting the reliability and validity of this safety climate scale. Given the widespread use of the Spanish language across the globe and that as of 2020 17.6% percent of the United States working population is Hispanic (BLS, 2021), there is a need for valid safety climate scales written in Spanish.
Methods
The translation-back translation procedure described by Brislin (1970) was followed in the current study involving three translators. All three translators (A, B, C) were bilingual researchers in the field of organizational behavior. In step one, translator A translated the English version of the survey into Spanish. In step two, translator B checked the Spanish translation and reconciled concerns with the original translator. The resulting Spanish survey was examined in cognitive interviews with Spanish-speaking workers, recruited from a single construction company. The purpose of these interviews was to examine the meaning of the translated items (for clarification purposes), and to identify language and/or content issues (e.g. indications of confusion). This procedure aimed to improve content validity and face validity. After the cognitive interviews, translators A and B discussed translation issues again and made additional changes. The resulting Spanish survey went to translator C for back-translation. All the translators then compared the original and back-translated English versions to reach final consensus on the Spanish survey. Surveys were distributed to four companies from various industries (e.g., manufacturing, construction, and utility) which had both sufficient English- and Spanish-speaking frontline workers. Survey data were collected from 12,148 frontline employees. The total sample for the English safety climate scale was 8,208 and for Spanish safety climate scale was 3,940. Both confirmatory factor analysis (CFA) and item response theory (IRT) approaches were adopted to test measurement equivalence to ensure the comparability of safety climate scores obtained from the English- and Spanish-language versions of safety climate measures.
Findings
Results from CFA indicated that, although item residuals were heterogeneous between the two versions of the safety climate scales, strong support was found for measurement equivalence with factorial invariance as well as invariant thresholds for item ratings. Results from IRT-based differential item functioning (DIF) analysis also supported strong measurement equivalence of the two versions of the safety climate scales.
Discussion
This study makes two important contributions. First, we provide a Spanish version of Zohar and Luria’s (2005) popular safety climate scale which has gone through a rigorous translation-back translation process. Second, we assure that the Spanish version of this scale is equivalent to the English version in terms of various psychometric properties. This helps to meet the demand of practitioners wishing to do more assessment work in organizations with Spanish-speaking employees. Further, having an equivalent Spanish version of the 2005 Zohar and Luria scale will support practitioners’ confidence that they are assessing the same construct regardless of the version of the scale they are using. As scholars, we must be certain that different language versions of a scale are equivalent for research exploring issues such as ethnic disparities in occupational safety and health. This study demonstrates that a test of measurement equivalency can provide confidence of the translation process from one language to another.
Conclusions
Conclusions: This study provides a much-needed safety-climate assessment tool for Spanish-speaking workers and paves the way for future research investigating factors related to safety climate for a growing Spanish-speaking workforce.
Reference: Brislin, R. W. (1970). Back-translation for cross-cultural research. Journal of Cross-Cultural Psychology, 1, 185-216. doi:10.1177/135910457000100301 Bureau of Labor Statistics. (2021, January 22). Labor Force Statistics from the Current Population Survey. Retrieved May 29, 2021, from https://www.bls.gov/cps/cpsaat11.htm Zohar, D., & Luria, G. (2005). A multilevel model of safety climate: cross-level relationships between organization and group-level climates. Journal of Applied Psychology, 90, 616-628.