Never Change a Winning Routine? Experimental Evidence on Performance Feedback Effects on Routines

This study explores the effects of performance feedback on organizational routines. In a laboratory experiment (N=184) with a 2 (performance level relative to aspirations: high vs. low) × 3 (feedback timing: early vs. continuous vs. late) design and a separate control condition, in which no feedback was provided, we analyze how routine change in the Cohen and Bacdayan (1994) card game is guided by reported performance levels and feedback timing. We differentiate between routine replacements and adjustments and show that both types of change are intertwined in that newly formed routines draw from previously established behavioral patterns. We find that feedback suggesting performance below aspirations is related to routine replacement, whereas we find no association between feedback above aspirations and routine replacement. Using the Cohen and Bacdayan (1994) measures for routinization as indicators for routine adjustment, we find that positive as well as negative feedback relates to decreases in speed an...