The accuracy and relevancy of the probit analysis for "in vivo dose-response laser tissue experiments"

The most common method of analysis for "in vivo laser tissue experiments" is the probit regression. The data gathered at these experiments are specific in that there are very few repetitions of the exact stimulus exposure; thus the response frequencies for most stimulus are either '0' or '1'. Though such type of data is acceptable in probit, it seems that such data might not produce robust estimates of the ED₅₀ and the slope. The accuracy of the probit's estimation was investigated by the use of Monte-Carlo simulation. Preliminary results suggest that the accuracy of the probit's estimations is conditional and might be biased in a way that raise doubts about the validity of the conclusions based on probit's estimations.