A case is made that Sternberg's additive factor logic may be particularly useful in examining the nature and locus of the effects of alcohol, fatigue and other factors on human performance. It is argued that an operation-predicated approach will prove more fruitful and general than the usual task-predicated approaches to alcohol and fatigue studies. This approach was exemplified by examining in detail data from extended practice in three memory search paradigms: Sternberg's fixed set and varied set tasks and Forrin and Morin's concurrent sets paradigm. While alcohol appeared to affect only the non-scanning components of the varied set task, more complex effects were noted in the other two paradigms. Fatigue and task complexity appeared to be important considerations in this regard. (A)
Abstract