Logarithmic least squares fuzzy model for pairwise comparison of nurse performance evaluation

We proposed a "basic model" logarithmic least squares fuzzy model for the pairwise comparison of nurse performance evaluation to reflect the effects of an evaluator's personal preferences and the rate scale used. In this study, we propose a new tranquility model for determining the evaluation elemental membership value estimated from the basic model. First, it involves the gathering of basic checklist data of nurse capability and paired comparison data from a single hospital in Fukushima Prefecture, from which we estimate the essential parameters (each evaluation element's weight w_j, evaluation phase k's membership value f_k, overall evaluated value Z_i) using the basic model. Second, we define F_ri as k's membership value in the overall evaluation phase and A_ri as the overall evaluation value of the analytic hierarchy process. We then calculate the Kullback-Leibler divergence between F_ri and A_ri and between Z_i and A_ri to consider the adequacy of the basic model. Moreover, we estimate the priority of nurse capabilities that should be emphasized using MRTQ, MTQ-1, and MTQ-2 in development of suggestions made in our previous work to consider the adequacy of our model.