THE CUTOFF PROBABILITY AS THE MEASURE OF QUANTAL IMPROBABILITY IN THEDILUTION METHOD

Citation
T. Kohno et S. Fukunaga, THE CUTOFF PROBABILITY AS THE MEASURE OF QUANTAL IMPROBABILITY IN THEDILUTION METHOD, Water research (Oxford), 32(10), 1998, pp. 3099-3107
Citations number
14
Categorie Soggetti
Engineering, Civil","Environmental Sciences","Water Resources
Journal title
ISSN journal
00431354
Volume
32
Issue
10
Year of publication
1998
Pages
3099 - 3107
Database
ISI
SICI code
0043-1354(1998)32:10<3099:TCPATM>2.0.ZU;2-P
Abstract
The MPN has been traditionally calculated on the basis of MPN tables o f two different dilution schemes and thus the bacterial density estima tion has been restricted to follow either of the above schemes. Howeve r, the use of a computer program may allow the use of any dilution sch eme and at the same time make it unnecessary to ambiguously truncate t he output result to three. On the other hand, it is well known that th e MPN can be calculated with any output quantum irrespective of how im probable the quantum may be. Thus, the improbability should be estimat ed also in the computer work. In the present study, the direct use of ''cut-off probability'' as the measure of improbability is newly propo sed on the basis that any quantum having maximum likelihood less than the probability can be regarded as improbable. The cut-off probability at the traditional erroneous rejection level of at most 1% was comput ed and then formulated into an equation applicable with some common di lution schemes. Furthermore, a static measure deviance was clarified t o have limited availability and therefore be erroneous if used in gene ral through comparison of the measure with the cut-off probability. (C ) 1998 Elsevier Science Ltd. All rights reserved.