An Empirical Investigation of Post-Completion Error A Cognitive Perspective

Forgetting to retrieve your original after photocopying, forgetting to collect your card after a withdrawal from a cash machine, are examples of a specific type of omission error termed post-completion error (Byrne & Bovair, 1997). A post-completion error (PCE) is the omission of a “clean-up” step after the main goal of a task is fulfilled. The error phenomenon has the property of being infrequent but persistent; it does not occur very often and, yet, it continues to occur now and again. This thesis is an empirical investigation of PCE to examine factors that provoke or mitigate the error. The investigation consists of two series of experiments. The first series of experiments is an extension of Byrne & Bovair’s finding of the effect of high working memory demand on the increased occurrences of PCE. A novel paradigm was designed and adopted in the experiments; it was found that PCE also occurs in problem-solving tasks, which impose a high demand on working memory load. Results from the experiments also suggest that the use of static visual cues may reduce the error rate. The second series of experiments investigates the effect of interruption on PCE in a procedural task paradigm. Based on the activation-based goal memory model (Altmann & Trafton, 2002) predictions were made on the effect of interruption position and duration on the error. Results show that PCE is more likely to occur with interruption occurring just before the post-completion step. Interruption occurring earlier in the task has no effect on PCE rate; it was found to be the same as having no interruptions at all. Moreover, interruption as brief as 15 seconds was found to be disruptive enough to increase PCE rate. The same disruptive effect was also obtained for other non-PCEs. The scarcity and disparate nature of the existing theoretical approaches to PCE motivated a meta-theoretical analysis of PCE. The analysis has resulted in the identification of the major criteria required for an adequate account of PCE. Although a complete cognitive model of PCE is beyond the scope of the current thesis, the meta-theoretical analysis offers new insights into the understanding of PCE and aids future theoretical development. The current thesis constitutes a methodological advance in studying PCE. New factors that provoke or mitigate the occurrence of the error were identified through empirical investigations. New insights into the understanding of the error were also possible through a meta-theoretical analysis within a coherent theoretical structure. 3