Moacir Manoel Rodrigues Junior

Professor Assessment Using Multi-Criteria Decision Analysis☆

Abstract This paper presents results from the pilot project on teacher assessment of the Regional University of Blumenau. It contains discussions about the project was elaborated by the team of internal assessments of the university, discussing the dimensions, issues and quantitative analysis techniques and the construction of the FURB Professoŕs Quality Index (IFQD). The multi-criteria tools used were the consensus theory, information entropy and the TOPSIS multi-criteria decision analysis method. Two undergraduate courses were assessed totaling 56 professors were evaluated by a group of 1979 students.

APPLICATION OF THE FACTOR ANALYSIS ON DATA MINING OF THE PERFORMANCE SOCIAL TO COMPANIES THE SECTOR CONSUMPTION CYCLIC LISTED ON BM&FBOVESPA

The research objective is to select the indicators of social performance to companies from the sector consumption cyclic listed on the BM&FBovespa, by applying the factor analysis as a data mining tool. To achieve the purpose, we performed a descriptive, document and quantitative research. As population was a delimited company in the consumer cyclic sector with the greatest representation in the economy. The final sample consists of 43 companies. To analyze the social performance, we used the indicators derived from the SVA. Regarding the analysis applied the technique of factor analysis as a data mining tool. It is concluded that for the period of 2009 selected factors were Personnel Spending and Tributes Spending. For the period 2010 and 2011, the factors were the variables, Quotient Value Added, Rate of Change in Gross Added Value and Rate of Change in Total Gross Value Added. Full Paper PDF: http://www.contecsi.fea.usp.br/envio/11contecsi-files/papers/820/submission/director/820-2538-1-DR.pdf

DATA MINING THE IDENTIFICATION OF THE CHARACTERISTICS OF THE BOARD OF THAT VALUE THE CORPORATE PERFORMANCE

The Board of Directors is an important instrument of corporate governance where standout practices that mitigate the conflict of interest between shareholders and managers ofcompanies, so that it can influence the performance of organizations. In this way we soughtfrom the techniques of rough sets and discriminant analysis, to identify the characteristicsof the board who value the performance of family firms listed on the BM & FBovespa.Was delimited study sample of 130 companies with characteristics of family ownership.The tools used for data mining techniques were the Rough Sets Theory and DiscriminantAnalysis. The results indicated that the main practices of the board with the greatestpositive influence on performance, are the total number of board members and the numbersubstitute of board. In turn, the technique of discriminant analysis highlighted only thevariable of the number substitute of board as significant for prediction with a hitpercentage of approximately 34%. By TCA all variables were confirmed, however thepercentage of information quality was 30% which can be considered low.