A comparison of positive and negative alcohol expectancy and value and their multiplicative composite as predictors of post-treatment abstinence survivorship.

Within social learning theory, positive alcohol expectancies represent motivation to drink and negative expectancies, motivation to restrain. It is also recognized that a subjective evaluation of expectancies ought to moderate their impact, although the evidence for this in social drinkers is problematic. This paper addresses the speculation that the moderating effect will be more evident in clinical populations. This study shows that (i) both expectancy and value reliably, independently and equally predict clients' abstinence survivorship following discharge from a treatment programme (and that this is almost entirely confined to the negative rather than positive terms). When (ii) expectancy evaluations are processed against expectancy through multiplicative composites (i.e. expectancy x value), their predictive power is only equivalent to either expectancy or value on its own. However (iii) when the multiplicative composite is assessed following the statistical guidelines advocated by Evans (1991) (i.e. within the same model as its constituents, expectancy and value) the increase in outcome variance explained by its inclusion is negligible and casts doubt upon its use in alcohol research. This does not appear to apply to value, however, and its possible role in treatment is discussed.