ABSTRACT
Three 95% confidence multipliers are shown in Table 4, labeled CM1, CM2 and CM3. CM1 is the 95% confidence multiplier used by Bailey & Eynon (1986). CM2 and CM3 follow discussion in O'Reilly & LaMotte (1987). Where CM1 employs a standard normal critical value, CM2 uses a 'Student's t' critical value in order to properly adjust for the fact that the population variance of ln(LC50) was not known and was estimated from the available data. This is the correct 95% confidence multiplier when applied to one LC50 determination. However, in the way that EPA uses Bailey & Eynon's 95% confidence multiplier, 133, the LC50 for the test mud is to be compared to the LC50 in the Alternate Toxicity Request (EPA variance procedure to allow operators with special mud additive needs to apply for a lower permissible toxicity limit). In this case, the 95% confidence multiplier is to be applied to a ratio of LC50s (equivalently, a difference in the logarithms of the two LC 5 00. Consequently, the lovrithm of the 95% confidence multiplier needs to be multiplied by
. This results in CM3. In summary, CM1 is not really the right answer to any question, CM2 is the appropriate 95% confidence multiplier to use on one LC50, and CM3 is the appropriate multiplier to use when comparing two LC50s.
