EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

ALTERNATIVES TO THE CLASSICAL FREQUENTIST CONFIDENCE INTERVAL FOR DESCRIBING ZERO-INFLATED LEAF DISEASE SEVERITY

Jhessica Letícia Kirch, Brena Geliane Ferneda, Fernando Henrique Silva Garcia, Sonia Maria de Stefano Piedade, Idemauro Antonio Rodrigues de Lara

Brazilian Journal of Biometrics · 2022

DOI: 10.28951/bjb.v40i2.563 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Brazilian Journal of Biometrics, 2022. Verfasst von Jhessica Letícia Kirch und 4 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This paper presents the bootstrap percentile interval and the Bayesian credible interval as alternatives to the classical frequentist confidence interval for analysis of zero-inflated data. The indicated methods were applied to soybean downy mildew severity data obtained by stratified sampling in two municipalities in the state of São Paulo: Estiva Gerbi and Piracicaba. The amplitudes of the frequentist and bootstrap percentile confidence intervals were similar. For the Bayesian approach, the credible intervals of the posterior predictive distribution were considered using the zero-inflated beta distribution as likelihood. The credible intervals showed a wider range and included values in the upper bounds of the intervals greater than those observed in the data. We conclude that Bayesian inference is more complex, but allows incorporation of prior information regarding regional and seasonal aspects, contributing to better disease management in the field. When this information is not known, nonparametric bootstrap resampling is a simple alternative to construct intervals for zero-inflated data without assuming the distribution function.

Zitieren

APA

Kirch, J. L., Ferneda, B. G., Garcia, F. H. S., Piedade, S. M. D. S., & Lara, I. A. R. D. (2022). ALTERNATIVES TO THE CLASSICAL FREQUENTIST CONFIDENCE INTERVAL FOR DESCRIBING ZERO-INFLATED LEAF DISEASE SEVERITY. Brazilian Journal of Biometrics. https://doi.org/10.28951/bjb.v40i2.563

Vancouver

Kirch JL, Ferneda BG, Garcia FHS, Piedade SMDS, Lara IARD. ALTERNATIVES TO THE CLASSICAL FREQUENTIST CONFIDENCE INTERVAL FOR DESCRIBING ZERO-INFLATED LEAF DISEASE SEVERITY. Brazilian Journal of Biometrics. 2022. doi:10.28951/bjb.v40i2.563

BibTeX
@article{Kirch2022,
  author = {Kirch, Jhessica Letícia and Ferneda, Brena Geliane and Garcia, Fernando Henrique Silva and Piedade, Sonia Maria de Stefano and Lara, Idemauro Antonio Rodrigues de},
  title = {ALTERNATIVES TO THE CLASSICAL FREQUENTIST CONFIDENCE INTERVAL FOR DESCRIBING ZERO-INFLATED LEAF DISEASE SEVERITY},
  journal = {Brazilian Journal of Biometrics},
  year = {2022},
  doi = {10.28951/bjb.v40i2.563}
}
Kostenlos starten →