İnsan kaynakları performans değerlendirmesine çok ölçütlü karar verme yaklaşımı

During last ten years, great majority of all the papers about human resourcesmanagement and performance focused on the human resources as department in anenterprise, personnel selection and the organization?s performance, but not on theworkers or employees? performance evaluation ? PE. According to this, it wouldn?t bebold to say there is a gap in the employees? performances evaluation field. Of coursein a conventional way, there are methods and tools to do that task but we have to askourselves if these methods are objective enough, if they are completely able to compareeffectively the employees, how they manage to assess simultaneously the tangible andintangible factors and if they take the interdependencies / interactions in consideration.In this study, trying to handle the tangible and intangible factors simultaneously, it isspecially focused on the last question, i.e. to consider effectively the interdependencies /interactions occurring between those factors. Hence this will lead us to the effectiveperformance evaluation of the employees.The aim of the proposed model is, by developing and improving the writer?s previousworks, to present a study on the employees? performance evaluation, a field that has notbeen highly considered in the literature of MCDM, and to bring a clearly higher level ofobjectivity to the performance evaluation for the employees working for a same task.Hence the MCDM perspective, added in the previous work to the human resources fieldto be more efficient at handling the tangible and intangible data simultaneously, will beimproved by taking into account the dependence or interaction between criteria andalternatives.In MCDM methods, most of the time there is the assumption of ?independence of thecriteria / sub-criteria / alternatives?. Because of that, people usually have a tendencyto construct independent criteria or criteria that are supposed to be so, which ispractically erroneous in real life problems and will lead to errors in judgments. To beable to do such assumption, one must be sure that there is no possible relation or inother words interdependence between criteria or alternatives according to the data athand, because the final decision may seriously change with this assumption. In thisstudy, the dependence of the criteria will be considered and analyzed very carefully.Two methods, which will be presented in this study for the purpose of dealing with theinteractions, handle different case of dependencies or interactions between criteria.One of the methods, namely the Analytic Network Process ? ANP, is one of the mostpopular MCDM tools which is able to handle the interdependencies among criteria /alternatives and considers the dependencies like one criterion having an effect onanother such as the aerodynamic and the fuel consumption of a car in case of selectionof car problem. Or the example of a teacher and two criteria to evaluate hisperformance: the knowledge on the course he/she is teaching and the teaching ability.So the interdependence to analyze here could be the effect of basic knowledge on theability of teaching.The other method, namely the Choquet Integral ? CI, is a fuzzy integral and considersthe interactions between k out of n criteria of the problem, which is called the kadditivityproperty. The interaction value (positive or negative) of the k criteria underconsideration will determine if these criteria must be satisfied at the same time or onlyone of them satisfied is enough to be considered as a successful or good performingunit. For the sake of simplicity and the applicability to the real life situations aseffectively as possible, in this study, the 2-additive CI will be used. Hence, with theexplanation given above, a positive interaction between two criteria will show that bothof them have to be satisfied and a negative one will show that it?s enough that one ofthe criteria be satisfied by an alternative in order to be considered successful. In otherwords, these cases can be considered as ?conjunctive? or ?disjunctive? relationsbetween criteria. It has to be noted that in these kinds of interdependencies /interactions, the relations between criteria and alternatives are not mentioned. This isjust the relations between criteria.As, in both methods which will be presented in this study, the decision makers - DM(s)will give their preference orders most of the time, it can be a wise approach to apply theSocial Choice Functions ? SCFs to the alternatives rankings when the situation will bea group decision. Therefore, for group decision cases, it is proposed to use anappropriate SCF, once the ranking of alternatives for each Decision Maker ? DM isfound. By doing this, a suitable ranking for the alternatives will be found in order tosatisfy the individually found rankings the most.The proposed performance evaluation method is as follows: First of all, the modelaccording to the organization?s requirements from its group of employees will be builtafter the meetings made with the DMs. The criteria and the sub-criteria will clearly bedefined. Then, in order to handle the interdependencies / interactions, the model willbe evaluated using ANP or CI or with a combination of these two according to thenature of the relations existing between the criteria, the sub-criteria and the alternativesof the model. Once the results for each DM are obtained, the SCFs will be used inorder to find the best suitable ranking of the employees to this group of DMs.