What signal is your inspection team sending to each other? Using a shared collaborative interface to improve shared cognition and implicit coordination in error-detection teams

Abstract

Coordination theory, created in a collaborative HCI context, has long emphasized the importance of implicit coordination. Such coordination is possible through interactivity design that creates shared cognition. Teams experience implicit coordination when they achieve tacit task coordination without direct or purposeful communication of task strategies among the participants, but this is impossible to achieve without shared cognition. Implicit coordination is often more effective than explicit coordination, and teams increasingly lack sufficient time or capabilities to coordinate explicitly; therefore, methods to foster implicit coordination could result in highly practical techniques to improve task performance. Implicit coordination, despite its promise, has not received much attention in the HCI literature. A particular problem we believe has hindered the application of implicit coordination is the dearth of objective measurement of implicit coordination and the related construct shared cognition. Teams have difficulty improving, implementing, and proving the value of implicit coordination without measurement. For our theoretical foundation, we use implicit coordination theory (ICT) to explain and predict how a simple shared interface in a group error-detection task can establish formalized group memory and group awareness, which foster the shared cognition necessary for implicit coordination to occur. To test ICT, we demonstrate innovative measurement of implicit coordination and shared cognition - without intrusive perceptual survey items - by leveraging measures from signal detection theory (SDT). We use the context of teams performing heuristic data-flow diagramming (HDFD) (which is similar to the Prepare, Collect, and Repair (PCR)-based process of heuristic evaluation) to measure the improved shared cognition and implicit coordination. Namely, we measured shared cognition with SDT's measure of response bias (c); we measured implicit coordination with SDT's measure of discrimination (d′). Supporting our ICT-based predictions, groups operating with a shared interface indeed experienced more shared cognition and implicit coordination - resulting in higher discrimination, more positive response bias, and more accurate hit rates and false-alarm rates than the control groups. Our results show great promise in making similar predictions, measurements, and improvements in other individual or group inspection tasks, which are prevalent in HCI development research and practice. © 2012 Elsevier Ltd.