Christian Köster presents various laboratory experiments on the issue of order quantity for non-storable goods in situations of uncertain demand within newsvendor decision contexts with asymmetric demand distributions. The author demonstrates that an adjusted, adaptive learning model has the highest explanatory power. This learning model can be utilized in practice as a forecasting model, and it becomes evident that experience in economic decision-making situations does not necessarily lead to better decisions. Continuous training and suitable decision support systems appear to be essential.