Utilization of the quantitative component of positive and negative results of short-term tests.

In this paper we discuss the possibility of utilizing not only the qualitative component of the information obtained from long-term and short-term tests (as is customary), but also the quantitative component of the results. We suggest that there is probably a precise mathematical relationship between the qualitative and quantitative approaches. We show that utilizing the quantitative approach, it is possible to give confidence limits to a given prediction, a possibility potentially very useful for risk evaluation. We show that starting from a reasonable working hypothesis, it is possible to include even negative data in a unified quantitative approach. Incorporating the quantitative component of the information could offer appreciable gains in predictivity, especially when utilizing batteries of tests.