Avaliando as escolas pelas suas competências: aplicação do DEA não discricionário para o ensino fundamental do Rio Grande do Norte

Formal education, understood by the gradual process that occurs at school, aims at learning and systematic knowledge is of great interest to society as it benefits its individuals and leads to many positive effects, such as increased productivity and welfare (Johnes, Johnes, 2007). Understanding what influences the educational outcome is as important as the result itself, because lets you manage these variables in order to obtain a better student performance. This work uses the data envelopment analysis (DEA) to compare the efficiency of Rio Grande do Norte schools. In this nonparametric method, an efficiency frontier was construct from the best schools that use the inputs set to generate educational products. Therefore, the data used were obtain by Test Brazil and year 2011 School Census to state and municipal schools of Rio Grande do Norte. Some of the variables considered as inputs and outputs have been obtain directly these bases - the other two were prepared, using the Item Response Theory (IRT) - they are the socioeconomic and school infrastructure indices. As a first step, we compared several DEA models, with changes of input variables. Then was chose the non-discretionary model for which was deep the analysis of results. The results showed that only seven schools were efficient in the 5th and 9th grades simultaneously; there were no significant differences between the efficiency of municipal and state schools; and there were no differences between large and small schools. Analyzing the municipalities, Mossoró excelled in both years with the highest proportion of efficient schools. Finally, the study suggests that using the projections provided by the DEA method, the most inefficient schools would be able to achieve the goal IDEB in 2011, in other words, it is possible to improve the education of significant state taking the efficient schools as a basis for too much.

Tradução e adaptação da escala de dependência de substâncias do Millon Clinical Multiaxial III para o Brasil

Millon describes the normal personality by means of adaptation styles that are effective in normal environments and personality disorders such as unadapted operating styles. To operacionalize his theoretical model, Millon has built several instruments, including the Millon Clinical Multiaxial Inventory III (MCMI-III), wich consists of a self report inventory composed by 175 true or false response items, containing four verification scales, and others scales wich evaluates 14 personality patterns and 10 clinical syndromes. The Substance Dependence scale (T) is placed along with Clinical Syndromes scales. This research is justified by the lack of a Brazilian instrument to assess personality psychopathological aspects, and aims to translate and semantically adapt the MCMI-III to the Brazilian context, checking validity elements of the Substance Dependence scale, and developing a computer application for assisting the evaluation of assessment results. To this intent, 2.588 individuals data was collected, male and female, aged between 18 and 85 years, characterized as belonging to a clinical or non-clinical group, who took part in the survey via the internet or in person. Respondents completed the MCMI-III, a socio-demographic questionnaire and a subgroup also answered to the Goldberg General Health Questionnaire (GHQ). Besides descriptive statistics, we performed the analysis using the Student t test, principal components analysis and internal consistency. Despite difficulties related to translating very specific English terms, the assessment by judges, experts on Millon´s theory, and the back translation, attested the adequacy of the Brazilian version. Factorial analysis indicated the grouping of translated T scale items into three factors (social activities prejudice, lack of impulse control, and oppositional behavior), by presenting a single item on a fourth factor (apparently related to seeking pleasurable stimuli). The Cronbach alpha for this set of items was 0,82, indicating an acceptable scale reliability. The data analysis resulted in distinction of scores between clinical and non-clinical groups and between men and women; the relationship between high scores on the scale T and the other scales; scores of drug users according to the declared used substance; and the relationship between high scores on T and the verification of disorder or risk on GHQ mental health factor, indicating the instrument´s adequate sensistivity in identifying psychopathologies and the relationship between the different disorders or psychopathological personality patterns. Although further studies are necessary to develop the scores transformation factors, the computerized correction tool was adequate.