Three experiments were conducted to assess the automaticity of event frequency processing. Using a modified concept-learning task, Experiment 1 showed that intentional frequency processing led to more accurate frequency judgments than incidental processing. Experiment 2 demonstrated that nonspecific (general memory) instructions in incidental processing conditions can actually lead to subjects' intentional processing of frequency information, which undermines the effectiveness of an intentionality manipulation. And, in Experiment 3, frequency processing accuracy was found to be interfered with by concurrent cover task capacity requirements, even though frequency processing occurred incidentally. The findings that frequency judgment is influenced by intentionality and by concurrent task factors clearly violate two of Hasher and Zacks ' (1979,1984) empirical criteria used to define automatic processing; they also challenge the assumption that automatic processing is always optimal. In light of our and others' data, either event frequency, the prototypical automatic process, is not automatic, or the assumption that a process must be optimal if it is to be considered automatic must be dropped. We suggest a less extreme view of automaticity such that nonconscious processes requiring minimal cognitive resources would be considered automatic, even if performance is not optimal.

No abstract

No abstract