Many theorists propose two types of processing: heuristic and analytic. In conflict tasks, in which these processing types lead to opposing responses, giving the analytic response may require both detection and resolution of the conflict. The ratio bias task, in which people tend to treat larger numbered ratios (e.g., 20/100) as indicating a higher likelihood of winning than do equivalent smaller numbered ratios (e.g., 2/10), is considered to induce such a conflict. Experiment 1 showed response time differences associated with conflict detection, resolution, and the amount of conflict induced. The conflict detection and resolution effects were replicated in Experiment 2 and were not affected by decreasing the influence of the heuristic response or decreasing the capacity to make the analytic response. The results are consistent with dual-process accounts, but a single-process account in which quantitative, rather than qualitative, differences in processing are assumed fares equally well in explaining the data.