Componentes principais na seleção de animais em pesquisas experiementais

• Autor(a) principal: Pinto, Renan Mercuri [UNESP]
• Data de Publicação: 2014
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://hdl.handle.net/11449/110369
• topic: Analise multivariada; Pesquisa experimental; Animais - Experimentação; Animais de laboratorio; Animals in research

Due to lack of statistical knowledge, many researchers use empirical or subjective methods to make a decision, ignoring the randomization process and thus influencing its results, particularly in the process of samples homogenization, essential for randomization in research involving animals as experimental unit. In this process, researchers have the erroneous habit of doing it in a fragmented or intentional manner, using only one convenient parameter to classify the group as homogeneous. Fact that besides resulting in inadequate homogenization, favors the possibility of discarding animals by simple spurious value than by bialogical similarity or dissimilarity. Considering this problem, this work suggests a multivariate statistical criterion of high accuracy developed in order to rescind the biological paradigm that evaluates the animal (experimental unit) in a fragmented manner. In this criterion, it is considered the animal biological condition as a whole, since the body reacts conjointly to any intervention ar treatment, in other words, all the parameters of the animal are analyzed together, as they are correlated, and the best way to understand animal behavior is expressed in a set of numerical information that faithfully represents its biological characteristics. This model was applied to a data set of animais with aortic stenosis and allowed the identification of six animais spurious to the group. It's remarkable contribution is explained by the improvement in the quality of homogenization and by ensuring with maximum accuracy the inclusion of similar animais and small frequency discards, motivating to maximize the homogeneous batch for subsequent submission to the experimental design by simple randomization process of the treatments.