Modelo misto via mínimos quadrados para delineamentos de blends de cafés

• Ano de defesa: 2018
• Autor(a) principal: Paulino, Allana Lívia Beserra
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Lavras; Programa de Pós-Graduação em Experimentação Agropecuária; UFLA; brasil; Departamento de Ciências Exatas
• Palavras-chave em Português: Modelos mistos; Blends de café - Análise sensorial; Delineamentos de misturas; Variável de processos; Mixed models; Mínimos quadrados; Least squares; Coffee blends - Sensory analysis; Mix designs; Process variable; Probabilidade e Estatística
• Link de acesso: http://repositorio.ufla.br/jspui/handle/1/28978

The present study aims to propose a mixed model in a sensory analysis of four blends of special coffea Coffea Arabica experiments, represented by Yellow Bourbon and Acaiá genotypes and non-special ones defined by Canephora and Comercial coffees. Each experiment was differentiated according to the processing type: dry and wet, and beverage concentrations defined as 0.07 and 0.10 (m / v), in which these amounts represented grams of coffee powder per 100 ml of water. In all experiments, the response variables were characterized by the sensory attributes: Taste, Bitterness and Note. For the implementation of the model, we considered the approach given by the least squares method, because it is easy to understand and implement to the researcher. The scripts were developed in the R software and made available in the Blendstat package. It was concluded that the aggregation of the experiments random effects contributed to the discrimination of the experiments, in which the blends of special coffees with the same origin (altitude), genotype and concentration were similar for the flavor attribute in both forms of dry and humid via processing.