Ordinal regression model for pea seed mass

Abstract

The development of seeds at various positions in the pod is asynchronous. Thus, the differences of seed dry mass production because of environmental conditions may depend on the cultivar type, type of inoculants and interrelations between seeds per pod, pods per plant or seeds per plant. Presently, a mathematical description of pea seed categorisation is missing. The aim of the study was the assessment of two groups of variables (quantitative and qualitative) for pea seed weight categorisation by ordinal regression model. Year, cultivar and inoculant constituted the first group (qualitative variables), whilst seeds per pod, the pods per plant and seeds per plant (quantitative variables) were entered as covariates in the ordinal regression model. According to the ordinal regression model variables, seeds per pod, pods per plant, seeds per plant, year and cultivar are meaningful predictors of the seed mass categories. However, the variable inoculant is marginally significant.

Keywords

• regression model
• seed categories
• inoculant
• legume
• pisum sativum
• logistisches regressionsmodell
• samengewicht inokulum
• legumiose
• pisum sativum