The Use of Cumulative Logit Models for Investigating the Inhibition of the Ascophaera Apis Fungus Fiber???s Growth Data

Abstract

The cumulative logit models are applied to the data of the extracted Chaemthad,
Tongpanchang and Sabsao which are the herbals used to inhibit the growth of a fungus fiber,
???Ascophaera apis??, of Patoomthanee and Phechaboon. The data obtained from the experiments of
the department of biology, faculty of Science, Silpakorn University. The objectives of this research
are (i) to study the effect of explanatory variables on the response variables, of which the high
capability level is more concentrated to inhibit the growth of a fungus fiber and (ii) to investigate the
cumulative response category probabilities. The results revealed that the statistical significant
factors, effecting to the response variable in high level, are the concentration (conc = 1), the
extracted Tongpanchang, the extracted Sabsao and the Butanol solvent. The estimated probabilities
under concentration
(conc = 1), the extracted Tongpanchang (ptype = 1), the extracted Sabsao (ptype = 2) and Butanol
solvent (solvent = 3) for high capability level of the response is approximately equal to 0.6639 and
those for high to medium capability levels of the response are approximately equal to 0.9635. These
probabilities are possibly and adequately high to inhibit the growth of the fungus fiber as well.

Keywords: Cumulative logit models, Proportional odds model, Ordered categorical variable, Ascophaera apis.