Replication Data for: Information and its Presentation: Treatment Effects in Low-information vs. High-information Experiments

This study examined how subjects react to a treatment under different information conditions. We recruited subjects from Amazon’s Mechanical Turk to follow a link (above) to a study where they would learn about two political candidates running for a congressional seat. As subjects entered the experiment they were randomly sorted into one of four conditions, each presenting the candidates in different ways based upon the amount of information available about the candidates and how it was presented. The four options were: • A short vignette about each candidate, including 5 attributes (Low-information, Static) • A static information board containing 20 attributes (High-information, Static) • A dynamic information board with 5 attributes (Low-information, Dynamic) • A dynamic information board with 20 attributes (High-information, Dynamic) Our treatment was a gender manipulation. In each condition, we determined which political party each subject preferred (Democrat or Republican) and then randomly assigned each subject to see their In-Party candidate appear as either a man or woman. The Out-Party candidate was always a man.