Collapsing Ordered Categories for Detecting the Emergence of Toxic Chemical by Using Dose-Response Curves

• Activity Number: 387
• Type: Invited
• Date/Time: Tuesday, August 3, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Memorial
• Author(s): Takashi Yanagawa*+
• Companies: Kurume University
• Keywords: categorical data ; dose-response ; monitor the safety of water

Abstract

This paper is concerned with constructing monitoring system based on the following experiment for detecting the emergence of toxic chemical in a fresh water system by using the behavior of snails. First, the behavior of snails when exposed to toxic chemical is classified into K ordered categories, reflecting severity of exposure. Next, snails are classified randomly into R dose groups, and snails in each group are exposed to pre-assigned dose of chemical for fixed hours in a day, and it is repeated for D days. The data may be summarized in a D x R x K table. For constructing simple and quick monitoring system it is desired to collapse K categories into two. Using dose-response curves we develop statistical method to achieve this goal so that these two categories may give the most sensitive detection for the emergence of the chemical.