COMPSTAT 2022: Start Registration
View Submission - COMPSTAT2022
Title: Sampling redesign considering spatial $t$-Student models: An effective sample size application Authors:  Luciana Pagliosa Carvalho Guedes - Universidade Estadual do Oeste do Parana (Brazil) [presenting]
Leticia Ellen Dal Canton - Universidade Estadual do Oeste do Parana (Brazil)
Miguel Angel Uribe-Opazo - Universidade Estadual do Oeste do Paraná (Brazil)
Tamara Cantu Maltauro - Universidade Estadual do Oeste do Parana (Brazil)
Rosangela Aparecida Botinha Assumpcao - Universidade Tecnologica Federal do Parana (Brazil)
Abstract: The spatial dependence modeling of georeferenced variables, in the presence of outliers, should consider the use of a robust probability distribution, such as the Student's t-distribution. This distribution can reduce the influence of outliers in the spatial dependence structure. Also, the financial investment in data collection and laboratory analysis of soil samples is a relevant factor when mapping agricultural areas. In this context, the reduction of sample size can be performed by the univariate effective sample size methodology (ESS), assuming that the $t$-Student model represents the probability distribution. An application of the ESS to redesign a sample configuration in an agricultural area with 167.35 hectares cultivated with soybean is presented to analyze the spatial dependence of soil penetration resistance with outliers.