Abstract
Under the moving average-angle method, LD50''s are estimated by transforming observed proportional responses to angular variables, obtaining 3-or 5-point moving averages of these angles and then interpolating linearly between two successive doses whose average angles bracket 45 degrees. An exact confidence limit formula for the LD50 may be obtained in terms of the interpolation ratio, A, and a defined variable g, commonly used to evaluate the strength of the dosage-response relationship. Detailed tables of confidence limits are presented for values of A and g between zero and unity. Finally, a formula is given for testing the significance of the difference between two LD50 estimates.

This publication has 3 references indexed in Scilit: