**Study Description** In this study, participants are asked to form predictions about Congressional election outcomes. Participants make predictions about which conditions predict whether Republicans will win or lose seats in the House of Representatives in a given election. Participants form their predictions in one of three experimental conditions. In the *Explore Only* condition, participants are given information about eight past elections and asked to form their predictions based on those data. In the *Confirm Only* condition, participants make predictions without any data, then are allowed to revise their predictions after seeing data from the same eight elections. In the *Explore/Confirm* condition, participants see data from four of the elections and form initial predictions. They then see data from the remaining four elections and can revise their predictions accordingly. After this phase, all participants rate their confidence in their final predictions. They are then shown data from five other elections and asked to predict the outcomes of those elections for Republicans. We are interested in examining whether the participants differ in their confidence and accuracy based on which prediction process they used. We are also interested in examining whether participants differ in the types of predictions they made based on the process they used. All study materials are located under Files. There is also a sample analysis script in Files that describes the planned sample and exclusion criteria, as well as the planned analyses.