STAFF APPRAISAL USING A BINARY REGRESSION

Abstract

The paper shows the possibility of the efficient evaluation of candidates for positions with the help of the binary-regression. The absence of expertise in using math methods by personnel departments makes recruitment process modeling inefficient, so the results obtained via binary-regression is of great importance.The purpose of the research is to show the relationship between the data in CVs and the fact of passing the probation period by employees. The author had at his disposal data of candidates’ CVs provided by several HR- agencies to their clients. Some of employees had passed the probation, some of them had not passed. To carry out the research the author chose three types of models - logit, probit, gompit.To estimate the parameters and the quality of the constructed models the author wrote the code in Maple computer algebra system. According to the results of the research the best predicting model was chosen.