Evaluation judgments were affected by information order and not by subsequent unconscious versus conscious deliberation. In three experiments, we examined the influence of early positive information on final evaluations of four objects. Based on a task analysis, we predicted primacy effects in judgments in a sequential data acquisition task. Thinking periods following presentation were used to manipulate conscious or unconscious processing. In all three studies, we found no effects of thinking manipulations but instead found reliable order effects. We developed and tested an online judgment model on the basis of the belief updating model of Hogarth and Einhorn. The model accounted for large proportion of the individual level variability, and model comparison tests supported the presence of a primacy effect. 1 We note that in Strick et al. (2010), participants saw all the information not in random order across options, but the information was organized such that all aspects of one option were viewed before moving to another option. This is a very different information acquisition paradigm from the used by Note: Standard deviations are in parentheses. Rating scale is À10 (very negative) to +10 (very positive).Online Judgment Model 211 C. González Vallejo et al.