This problem is even more complicated with a behaviorally defined phenotype such as alcoholism, as there is no completely objective laboratory test for the phenotype. In this analysis, we explore the problem of phenotype definition with the Collaborative Study on the Genetics of Alcoholism (COGA) dataset, and look at alternate definitions of the phenotype. Of particular interest is the problem of lack of consistency in results due to a change in the definition of alcoholism from DSM-III-R to DSM-IV criteria [3].
