The coefficient of reliability is often estimated from a sample that includes few subjects. It is\ntherefore expected that the precision of this estimate would be low. Measures of precision such as\nbias and variance depend heavily on the assumption of normality, which may not be tenable in\npractice. Expressions for the bias and variance of the reliability coefficient in the one and two way\nrandom effects models using the multivariate Taylor�s expansion have been obtained under the\nassumption of normality of the score (Atenafu et al. [1]). In the present paper we derive analytic\nexpressions for the bias and variance, hence the mean square error when the measured responses\nare not normal under the one-way data layout. Similar expressions are derived in the case of the\ntwo-way data layout. We assess the effect of departure from normality on the sample size requirements\nand on the power of Wald�s test on specified hypotheses. We analyze two data sets, and\ndraw comparisons with results obtained via the Bootstrap methods. It was found that the estimated\nbias and variance based on the bootstrap method are quite close to those obtained by the\nfirst order approximation using the Taylor�s expansion. This is an indication that for the given data\nsets the approximations are quite adequate.
Loading....