In today's enterprises, forecasts of time series are a crucial part of the planning process. Human experts often create these forecasts -for example, in cash flow or sales forecasting. The human participation can lead to forecasts being influenced by cognitive biases like anchoring and adjustment. This study aims to detect anchoring and adjustment effects in forecasting processes based on the newly developed Bandwidthmodel. We show that the Bandwidthmodel has higher explanatory power with regard to the relation between anchoring and adjustment effects and forecast errors in comparison to other models based on synthetic forecast series. These series allow the generation of specific pattern of anchoring and adjustment effects. The results suggest that usage of the Bandwidthmodel can improve the accuracy of forecasts and is beneficially for a forecast support system.