The issue before the New Jersey Supreme Court in the Frye hearing New Jersey v. J.L.G. (2018) was whether the scientific community agreed that Summit's (1983) Child Sexual Abuse Accommodation Syndrome rested on a firm scientific foundation. Lyon et al. (this issue) critique our approach to describing child sexual abuse disclosure, which involved extrapolating rates from children who came to the attention of authorities. Lyon et al. claim that our conclusions are marred by sampling biases resulting from what they term the ground truth problem, suspicion bias and substantiation bias. The points Lyon et al. claim we "fell victim to" were the very points we acknowledge are inherent difficulties in estimating the extent to which children will come forward to tell others about sexual maltreatment. Lyon et al. offer an alternative solution to the inherent difficulties in studying a difficult-to-identify population, relying in large part on 21 papers published mostly in the 1960s and 1970s. We argue that the method they propose has more flaws than the one it is intended to replace. Points of agreement and disagreement, along with suggestions for future research, are discussed. Moving forward, we argue that studies are needed that embrace both validity and generalizability in order to foster data-driven theories rather than invoking the intuitive suppositions of Summit's (1983) syndromal evidence.
© 2020 John Wiley & Sons Ltd.