Decision making trial and evaluation laboratory method and semantic structure analysis are the two most popular approaches to study structural analysis of critical factors or dimensions but do not have solid theoretical backgrounds to construct causal relationships among critical factors or dimensions. This study proposes a framework by using normalized mutual information for both continuous and discrete random variables to set up the contextual relationships from the raw data. A Monte Carlo simulation based on normalized mutual information is conducted to estimate the threshold value of establishing the causal relationships. Two internal survey data of safety attitudes questionnaire in terms of thirty questions and six dimensions are used from viewpoints of physicians and nurses. Four cases are illustrated to show how the proposed framework works in practice to identify critical factors or dimensions from the internal survey data. This proposed framework enables hospital management to initiate improvements from causal factors or dimensions to effectively enhance patient safety from medical staffs’ viewpoints.
|Number of pages||8|
|Publication status||Published - 2019 Aug 12|
All Science Journal Classification (ASJC) codes