Figure 2

Relationship between for the CRisch, Odds and Probit models and , heritability on the underlying liability scale. (a-c) For each , τ is estimated from the Probit model simulation and used as an input for the other models, so that all three models are benchmarked by K and τ. The shape of the relationship is not dependent on the choice of n and p; the τ when = 0.1, 0.3, 0.5, 0.7 and 0.9 are listed above each graph when n = 1,000 and p = 0.3. From simulations of a single population of 10⁶ individuals.