“…Examples are threshold models such as the hockey stick model [3,15,16] and the hormesis model introduced by Hunt and Bowman [17], but these models are not in common use, which may be due to the lack of a biological rationale for the existence of a threshold. Second-order and higher-order polynomial models, which have occasionally been used for modeling dose-response data [4,18,19] and polynomials in dose within the generalized linear model framework [20,21], are also not part of the general framework proposed in the present study, but these models are only occasionally useful in a dose-response context, as even interpolation can be unreliable [6].…”