Numerical precision on judgmental forecasting

Abstract

Numerical precision refers to precision of a number defined by the number of ending zeroes. For example, 1,459 is comparatively precise number than 1,400

1,400 is comparatively round number than 1,459. We investigate the effect of numerical information format on judgmental forecasting task. In other words, for demand forecasting task, we investigate the impact of information related to numerical precision on the task. Furthermore, we examine the precision level of numbers used in forecasts as each numerical information format. We find that individuals will less chase the most recent period demand if they receive demand information with precise numbers and individuals will estimate the mean of demand well if they receive demand information with round numbers. Moreover, we reveal that the forecasting performance is lower if individuals make forecast using precise numbers.