Please use this identifier to cite or link to this item:
|Title:||ANÁLISE CONJUNTA BASEADA EM ESCOLHAS: UM ENFOQUE VIA MODELO LINEAR GENERALIZADO DE POISSON|
CHOICE-BASED CONJOINT ANALYSIS: AN APPROACH BY POISSON GENERALIZED LINEAR MODEL
|Publisher:||Editora UFLA - Universidade Federal de Lavras - UFLA|
|Description:||This paper presented a practical application of the choice-based conjoint analysis methodology. It was proposed the generalized linear Poisson model to analyze the data. The study was conducted with choices regarding the evaluation done by 144 consumers, eight samples of light yoghurt strawberry flavor, varying the information of three ingredients (sugar, fat and protein) in a full factorial design. Inferences and results are presented in terms of estimating the main effect of the attributes, probability and ratio of choice. It was found that the generalized linear Poisson model was equivalent to the multinomial logit model and can be used for the analysis in question. Practical results: the most likely choice was associated with yoghurt containing strawberry flavor light information: 0% of sugar, 0% fat and enriched with bioactive proteins.|
This work was presented an application of Choice-based Conjoint Analysis (CBCA) methodology, however, for data analysis, it was proposed the use of linear Poisson Generalized model. The study was conducted with data of choice for the assessment, by 144 consumers, of eight samples of light yogurt strawberry flavor, varying the information of three ingredients (sugar, fat and protein) in a full factorial design. Inferences and results were presented in terms of estimating the main effect of the attributes, probability and choice ratio. It was concluded that the Poisson model can be used for data analysis in the CBCA, by presented similar results to traditional analysis employing the multinomial logit model. The practical results, the most likely choice was associated with yoghurt containing strawberry flavor light information: 0% of sugar, 0% fat and enriched with bioactive proteins.
|Appears in Collections:||Revista Brasileira de Biometria|
Files in This Item:
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.