Koroto, Tadewos(PhD)Woldemariam, Solomon2018-06-282023-11-092018-06-282023-11-091999-05http://10.90.10.223:4000/handle/123456789/4419Many methods of low dose risk assessment based on the popularly used dose response models have been suggested. The estimates of risk by the different methods were observed to differ largely. In this paper, two alternative approaches of low dose risk assessment are suggested. Both methods are based on the logistic dose response model. The underlying principle in both approaches is to make the lower tail of the logistic curve heavier. The performances of the methods are compared against standard logistic estimates and linear extrapolation estimates based on logistic model by the Monte Carlo method. The results indicate that the performance of one of the new approaches is superior when the true dose response model is Multihit or Logistic. Linear extrapolation is the best method when the underlying model is Onehit, Multistage or Weibull. The performances of the estimators are obselVed to depend on the choice of number of animals per dose group and risk level to be estimatedenExtrapolation Based on LogisticThesis