Heuristics, biases and improvement strategies in judgmental time series forecasting

There is evidence that forecasts produced in business and other organizations often involve substantial elements of human judgment. In forming their judgments forecasters may have access to either time series or time series and contextual information. This paper reviews the literature to ascertain, for each information level, what is currently known about the heuristics people use when producing judgmental point forecasts and the biases which emanate from the use of these heuristics. The paper then uses evidence from the literature to review the effectiveness of a number of strategies which are designed to improve the accuracy of judgmental point forecasts.