Improving Forecasting Accuracy by Combining Statistical and
Judgmental Forecasts in Tourism.
Section V considers robustness analysis and extensions, showing in particular that
judgmental forecasts have adjusted faster than the others to capture developments during the Great Recession.
In a later work Sanders and Ritzman (2004), distinguishes between the marketing function, which more typically generates
judgmental forecasts, and operations, which rely more heavily on quantitative data, and suggests integration techniques for these approaches.
Also, Makridakis (1987) indicated that managers should prepare a separate
judgmental forecast, and then objectively combine it with quantitative forecasts, rather than use judgment for the blending process itself.
Cognitive biases do much to explain the limitations of
judgmental forecasts, but Tetlock's work suggests an additional explanation.
Future research related to the PMI should explore the use of combining
judgmental forecasts with the quantitative forecasting of the PMI.
The main claims Sims makes are, first, that the Federal Reserve forecasts well, especially when forecasting inflation; second, that the informational contents of different forecasts are highly correlated, so that strong claims of superiority of one forecast over another should be treated as suspect; and third, that there does not appear to be strong evidence that the
judgmental forecasts of the Federal Reserve are superior (as measured by the root mean square forecast error) to its model-based forecasts.
One difficulty with
judgmental forecasts, however, is that it is hard, if not impossible, for an outside observer to trace the source of systematic forecast errors because there is no formal model of how the data were used.
Much of this work has concentrated on forecasts produced by various time series methods of extrapolation for individual series, although there have also been other studies comparing econometric and/or
judgmental forecasts with the consensus.
Averaging can also work with purely
judgmental forecasts. The error associated with these forecasts, much less averages of them, have not been studied in nearly as much depth as quantitative forecasts, so the guidance on how many forecasts to average is not as definitive.
The advantages and limitations of economic models and
judgmental forecasts are reviewed, and a process that incorporates features of both is recommended.