Evidence for an Artifactual Interpretation

Proponents of unconscious-thought theory assert that letting the unconscious ‘‘mull it over’’ can enhance decisions. In a series of recent studies, researchers demonstrated that participants whose attention was focused on solving a complex problem (i.e., those using conscious thought) made poorer choices, decisions, and judgments than participants whose attention was distracted from the problem (i.e., those purportedly using unconscious thought). We argue that this finding, rather than establishing the existence of a deliberation-withoutattention effect, is explained more compellingly in terms of the well-established distinction between on-line and memory-based judgments. In Experiment 1, we reversed the recent finding by simply changing participants’ on-line processing goal from impression formation to memorization. Experiment 2 provided a replication and further established that some cognitive effort appears necessary to produce both the original pattern of results and its reversal, suggesting that such judgments are ultimately a product of conscious, rather than unconscious, thinking. Martin Luther King, Jr., astutely noted, ‘‘Nothing pains some people more than having to think’’ (1963/1981, p. 14). Unconscious-thought theory (UTT; Dijksterhuis & Nordgren, 2006) appears to offer solace to such individuals. According to UTT, when it comes to making complex decisions, conscious reasoning requiring cognitive effort is maladaptive; ‘‘unconscious thinking,’’ defined as effortless deliberation in the absence of attention directed at the problem domain, produces superior judgments. The evidence purportedly supporting this deliberation-without-attention effect derives from a paradigm that assumes that unconscious thought predominates when attention is diverted from a decision task via distraction (Dijksterhuis, 2004; Dijksterhuis, Bos, Nordgren, & van Baaren, 2006). However, we propose that the deliberation-withoutattention effect is an artifact that can be explained more simply in terms of on-line and memory-based judgments (Hastie & Park, 1986).