Managerial Implications of Extended Breaks on Predicting Performance Times for Industrial Tasks

Abstract

Production Managers and Industrial Engineers have relied upon learning (progress) curves for over fifty years. However, until recently only the impacts of extended breaks on performance time predictions were considered. This study examined the effect of breaks on two typical simulated industrial tasks. Fifty-eight subjects performed either the traditional peg-board, a low cognitive task, or a spreadsheet graphic, a moderately-high cognitive task, for 28 iterations. Upon completion of the assigned task, a break period that ranged from 2 to 83 days was randomly assigned to each subject. After the break, subjects replicated their assigned task. Regression analysis was used to select the best model to predict the performance time for the first iteration after a break. An exponential model was selected for the low cognitive task and a multiple linear model for the moderately-high cognitive task. Both models selected were no-intercept models and had multiple correlation coefficients of 0.729 and 0.897 respectively. The ability to accurately predict the first iteration time after a break is a key element in calculating time lost to forgetting and determining the forgetting function. These models may be useful in assisting production managers and industrial engineers in establishing more realistic progress curves and accurate standard times, thus reducing excessive idle time.