lc50_probit: LC Estimation by Probit Analysis (Maximum Likelihood)

In insectecol: Insect Ecology Data Analysis Toolkit

LC Estimation by Probit Analysis (Maximum Likelihood)

Description

Estimates the lethal concentration by maximum-likelihood probit analysis in the sense of Finney: a binomial generalized linear model with probit link fitted by iteratively reweighted least squares.

Usage

lc50_probit(d, lc = 0.5)

Arguments

d	Same as lc50_traditional.
lc	Same as lc50_traditional.

Details

Unlike the two regression methods, the line is fitted by maximum likelihood (a quasibinomial GLM with probit link, fitted by iteratively reweighted least squares) to the Abbott-corrected proportions p = (p_raw - p_c) / (1 - p_c) with the numbers tested as weights (the classic Finney effective-counts formulation), rather than by least squares to probit-transformed points. Groups with a corrected mortality of exactly 0 are dropped, as in the two regression methods. A quasibinomial family is used, so the covariance matrix of the coefficients incorporates the heterogeneity factor (Pearson chi-square divided by the residual degrees of freedom): the confidence intervals are automatically widened when the data show more variation than the binomial assumption allows. The reported chi-square statistic and its p value serve as a goodness-of-fit test of the probit-log concentration line.

Value

Same as lc50_traditional (with fit being the fitted glm object).

References

Finney, D. J. (1971) Probit Analysis, 3rd edition. Cambridge University Press, Cambridge.

See Also

lc50_traditional, lc50_improved

insectecol documentation built on Oct. 5, 2026, 5:08 p.m.