Implicit evaluation of familiar and novel concepts presented at low levels of conscious detectability.

In 2 experiments, implicit evaluation of novel and familiar concepts was assessed using a sequential priming procedure that enabled estimates of evaluative priming effects at low levels of detectability. In Experiment 1, the novel concepts referenced common names, and in Experiment 2 they referenced nonsense words. Whereas familiar concepts yielded priming effects at low levels of detectability in both experiments, novel concepts did not elicit any priming effect. Implicit evaluation of novel concepts has been documented in related research but under conditions that differ from those investigated here. The present results identify important limiting conditions associated with the implicit evaluation effect.