Using standardization for fair data evaluation

Abstract

Evaluation on a specific criteria is divided into quantitative evaluation and qualitative evaluation, the former can clearly depict scores in objective manner while the latter is obscure to measure. To reduce the obscurity, we used formula of Z-score and T-score through 'standardization of normal distribution'. To testify the effect, we set a simulation of two groups of which consisted 3 interviewers and 5 interviewees. The result after using the method lead to equal mean and equal standard deviation regardless of different interviewers. Also, by using T-score, we again rearranged scores once again by using T-score, in order to enable comparing between criteria possible. By doing so, we summed up the altered scores, and arranged the ranks of interviewees from the highest to the lowest. This proved to be effective on reducing obscurity compared to raw scores' sum up. © 2018 Global IT Research Institute (GiRI).