In the recent decades, economic research has begun to intensively analyze the role of agents’ expectations for economic outcomes. Academic studies acknowledge that the link between a piece of information and subsequent market development is inevitably the perception of information by economic actors and their expectation formation and decision making. Normative decision theories assume that people have clear subjective beliefs, which are invariant with respect to the manner in which information is represented and the manner in which they are elicited. However, violations of these invariance assumptions, denoted as framing effect, are documented in an extensive body of literature. Within the scope of this project we will focus on the difference in stock market expectations and risk perception arising from different elicitation forms, as well as its determinants. For this purposes we design a natural field experiment as well as internet and lab experiments. Furthermore, the project will evaluate different elicitation methods in terms of forecast quality. Last but not least, we will analyze which elicitation method individuals choose endogenously in connection to different tasks (forecasting, investment decisions etc.).