Abstract
Background: Workplace bullying interferes with provision of optimal care to patients and contributes to decreased job satisfaction and withdrawal of nurses from the workforce. Little is known about bullying (ijime) or its measurement among Japanese hospital nurses.
Objectives: The objectives of this study were to describe responses and explore dimensionality of a Japanese translation of the 23-item revised Negative Acts Questionnaire (NAQ-R), a frequently used measure of bullying.
Method: Responses of 881 registered nurses working in hospitals in Japan who answered all 23 NAQ-R items were analyzed. Descriptive statistics were used to summarize item responses, the principal components analysis approaches used by other international investigators were replicated, and five exploratory maximum likelihood factor analysis models were estimated.
Results: Responses to the 5-point (1 = never to 5 = daily) NAQ-R items were skewed, and 19% of the nurses replied never to all 23 items. The principal components analysis produced three components with eigenvalues greater than 1, and all five maximum likelihood exploratory factor analytic models were rejected using the chi-square test statistic. Model comparison based on the Akaike Information Criterion identified the five-factor maximum likelihood model as the best approximating structure.
Discussion: Dimensionality of the NAQ-R item set included verbal bullying, physical bullying, exploitation, undervaluation, and isolation. The solution reflected experiences of bullying reported in international studies, unique characteristics of Japanese bullying, and skewness in the data. Item response theory is recommended as an alternative way to gain insight into item functioning when the NAQ or its translations are used to measure nursing workplace bullying.