Mixtures of normal distributions: modality and failure rate (Q2567183)

Consider a mixture of two normal distributions with the same variance, \(\sigma^2\), say. Let \(\mu_1\) and \(\mu_2\) be the corresponding means, and let \(p\) and \(1-p\) be the corresponding weights in the mixture, \(p\in[0,1]\). The authors show that if \(2\sigma<| \mu_1-\mu_2| \) then for all \(p\in[0,1]\) the mixture has increasing failure rate. They also study the behaviour of the failure rate function for other cases; that is, when \(2\sigma\geq| \mu_1-\mu_2| \), and the mixed distributions have unequal variances.