Analysis of covariance for retrospective bias control in agricultural research: an application using R

 

Guardat en:
Dades bibliogràfiques
Autors: Vargas Rojas, Jorge Claudio, González Calero, Meilyn, Piña Duarte, Marileny
Format: artículo original
Estat:Versión publicada
Data de publicació:2026
Descripció:This work was developed within the framework of the Experimental Design II course in the Agronomy program at the University of Costa Rica in 2023, aiming to explain the analysis of covariance (ANCOVA) as a data analysis method using the R programming language. A database describing an experiment under a completely randomized block design (CRBD) was employed, where the number of plants per experimental unit was measured to control its potential effect on the yield of six corn varieties. Four models were fitted: the model_rl to assess viability, the model_hp to assess the homogeneity of slopes, and two additional models, the model_scov without the covariate effect, and the model_cov with the covariate. For the evaluation of these last two models, penalized information criteria [Akaike information criterion (AIC), Bayesian information criterion (BIC)] and hypothesis testing based on the likelihood ratio test (LRT) were utilized. The model_cov exhibited a better fit than the model_scov; the inclusion of the covariate reduced residual error, resulting in greater precision in estimates and enabling an unbiased comparison of means. ANCOVA is a strategy to consider secondary sources of variation that are difficult to control but, due to their quantitative nature, are measurable.
Pais:Portal de Revistas UCR
Institution:Universidad de Costa Rica
Repositorio:Portal de Revistas UCR
Idioma:Español
OAI Identifier:oai:portal.revistas.ucr.ac.cr:article/6538
Accés en línia:https://revistas.ucr.ac.cr/index.php/rintersedes/article/view/6538
Paraula clau:Statistical models
Data analysis
Biostatistics
Analysis of variance
Statistical analysis
Covariance analysis
Modelos estadísticos
Análisis de datos
Bioestadística
Análisis de varianza
Análisis estadístico
Análisis de covarianza