Evaluating the influences of confounding variables on benchmark dose using a case study in the field of ionizing radiation