89

Studies have found that positive moods are associated with more optimistic likelihood estimates than are negative moods. However, it is still unclear whether these effects can be attributed to the direct influence of mood and to which extent they could be at least partially explained in terms of application of naïve theories. In experiment 1, we observed mood congruence-like effects among participants who were simply asked to apply their naïve theories regarding the effects of mood on likelihood estimates. Experiment 2 revealed that mood congruence effects (moods being manipulated) were obtained only when participants were not constrained to respond honestly. When participants gave their estimates under bogus pipeline, they no longer showed mood congruence effects. These findings suggest that experimental demand could be a plausible explanation for some mood congruence effects observed in the laboratory on likelihood judgments.