Inspired by Sunstein and Thaler's "Nudge" concept and Fogg's "Persuasive Technology", we ran a user study and found that subtle, non-functional changes in the interface of an analysis support module (top) can generate statistically significant changes in users’ analytic behavior of a visual problem solving task. A first set of changes nudged subjects to increase their use of the analysis module by 24% (lower left, p=0.02) in an attempt to support our subjects’ working memory. It also caused them to switch between hypotheses 35% more often (lower center, p=0.03), indicating more consideration of alternative hypotheses. A second set of changes then lead subjects to gather 26% more evidence per hypothesis (lower right, p=0.01).