are usually between 0.3 and 0.4, although figures can approach 0.6 when only individuals who are clearly high and low on LR are considered (Schuckit et al., 1997 a, b). Thus, while the correlations between alcohol challenge and SRE-based LR scores are significant, it is not clear what the less than perfect concordance means. An important question is whether the correlation indicates that these measures overlap in their ability to predict future heavy drinking and alcohol-related problems. Therefore, this paper presents the first report of the concordance in prediction of future heavy drinking in subjects who received both alcohol challenges and filled out the SRE at a similar age and participated in a subsequent follow-up.
