Predictive Distributions in the Presence of Measurement Errors

Suppose a characteristic E, varies throughout a population, and the probability is needed that E, exceeds a given threshold value, L. This paper covers the situation where there is measurement error, i.e., only x - ξ + ∊ can be observed, and it is assumed that both ξ and e have normal distributions. The paper treats two cases, (i) where the ratio of the variances of ξ, and ∊ is known, and (ii) where it is unknown. An example illustrates the results.