Modelos de equações estruturais com regressões adaptativas e construção de um índice para validação de construto com aplicações na discriminação de perfis de consumidores de cafés especiais
Year: 2021
Type: dissertation
Abstract: This work consists of presenting a new approach to Adaptive Linear Regression adapted to \nstructural equation models and improving the index related to the Average Variance Extracted \n(AVE), given a plug-in approach, and replacing the error variances with the factor loadings of \nthe estimated adaptive regressions. To do so, a Monte Carlo simulation study was performed \nconsidering scenarios with different numbers of outliers, which were generated by distributions \nwith symmetry deviations and kurtosis excess. Sample sizes were defined as n=50, 100 and \n200. In formative structural models and considering outliers generated either from symmetrical \ndistributions or from multivariate log-normal distributions, the Adaptive Linear Regression modeling \nwas found to be efficient in the different scenarios under analysis. Likewise, for models \nwith specification errors, this method was proven to have low efficiency, as expected. Furthermore, \nconstructs were elaborated with variables that could enable both the characterization and \nthe distinction of individuals among the different groups of Brazilian specialty coffee consumers \nand that could provide different perspectives on the transition among them. The results \nmade it possible to better distinguish the consumers and better characterize the proposed categories, \nthus contributing to the improvement and simplification of marketing strategies used by \nplayers in this market. In addition, the results also promoted the discussion on which factors \nstimulate the transition of an individual from an initial construct to another, and we showed that \ntransitioning from regular consumers to enthusiasts is easier than moving from enthusiasts to \nspecialists.
Author: Santos, Patricia Mendes dos
Institution: Universidade Federal de Lavras
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7015395212
