Reference dependence is a key concept in behavioral economics, but it is debated how the reference point is determined. We contribute to this debate by testing the predictions of different reference-dependent models (status quo- versus expectations-based) in the context of information preferences. In our experiment, subjects participate in two monetary lotteries and have to decide how they want to be informed about the outcome of these lotteries. For half of the subjects we frame the lotteries as two gain lotteries, for the other half as two loss
lotteries. Based on Thaler (1985) one can expect that people want to learn the outcomes of the gain lotteries separately and the outcomes of the loss lotteries clumped together (cf. hedonic editing hypothesis). On the other hand, the expectations-based reference dependent model of Koszegi and Rabin (2009) predicts that irrespective of the frame, subjects should prefer clumped information. The experimental data shows a preference for segregating information about gains. As for losses, clumped information and segregated information are equally popular.