Bootstrap inference for quadratic regression model confidence interval

This study was carried out with the purpose of proposing a construction of confidence intervals for the critical point of a second degree regression model using a parametric bootstrap methodology. To obtain the distribution of the critical point, height growth data of the plants were used. From the...

Full description

Bibliographic Details
Authors: Gouveia, Nicásio, Mateus, Ana Lúcia Souza Silva, Silva, Augusto Maciel da, Ferreira, Leandro, Aimi, Suelen Carpenedo
Format: article
Status:Published version
Publication Date:2020
Country:Brasil
Institution:Universidade Federal de Santa Maria (UFSM)
Repository:Revista Ciência e Natura (Online)
Language:Portuguese
OAI Identifier:oai:ojs.pkp.sfu.ca:article/40501
Online Access:https://periodicos.ufsm.br/cienciaenatura/article/view/40501
Access Level:Open access
Keyword:Parametric bootstrap
C. canjerana
Theoretical variance
Bootstrap paramétrico
Variância teórica
Description
Summary:This study was carried out with the purpose of proposing a construction of confidence intervals for the critical point of a second degree regression model using a parametric bootstrap methodology. To obtain the distribution of the critical point, height growth data of the plants were used. From the analysis, the theoretical variables for the error and the confidence intervals were constructed. In addition, we examined different variance expressions with the purpose of the bootstrap-t confidence interval. The point estimate of the critical point was 10.7423 g L-1 of fertilizer doses without growth of C. canjerana plants. It was verified that the confidence intervals that considered the expression of the variance with the covariance between the regression models, present more satisfactory results, that is, results with more precision.