A Theoretical Framework to Understand and Engineer Persuasive Interruptions

A Theoretical Framework to Understand and Engineer Persuasive Interruptions Muhammad Walji (Muhammad.F.Walji@uth.tmc.edu) University of Texas Health Science Center at Houston, School of Health Information Sciences, 7000 Fannin, Houston, TX 77030 USA Juliana Brixey (Juliana.J.Brixey@uth.tmc.edu) University of Texas Health Science Center at Houston, School of Health Information Sciences, 7000 Fannin, Houston, TX 77030 USA Kathy Johnson-Throop (Kathy.A.Johnson@jsc.nasa.gov) NASA Johnson Space Center, Houston, TX 77058 USA Jiajie Zhang (Jiajie.Zhang@uth.tmc.edu) University of Texas Health Science Center at Houston, School of Health Information Sciences, 7000 Fannin, Houston, TX 77030 USA positive outcomes, while at the same time minimizing some of their most disruptive properties. After all, interruptions are constantly used to help manage and complete important everyday tasks. Such interruptions also have the ability to influence and change behavior. In order to better understand and explain how interruptions can be engineered to be positive and persuasive we propose a theoretical framework and conceptualization. The theoretical framework may also guide designers on discovering factors to help develop appropriate interruptions. Abstract Interruptions are often seen as distracting or sometimes devastating elements that need to be minimized or eliminated. However, interruptions are also used to increase efficiency, productivity, prevent errors, and even influence behavior. Existing theories and taxonomies of interruptions fail to account for the helpful aspects of interruptions. Therefore we propose a theoretical framework to help explain the positive aspects of interruptions. Warnings & alerts, reminders, suggestions and notifications are examples of interruptions that have beneficial outcomes by changing and influencing behavior. We propose a cognitive theory of interruptions based on the properties of the users, their tasks, and best presentations depending on the desired effectiveness of the interruption. Norman’s 7-stage action model serves to explain how and why an interruption is accepted, and potential mismatches between the goal of the interruption and the user. Potential applications of this model include better understanding the effects of interruptions, and guidance to design effective and persuasive warnings and alerts, reminders, suggestions and notifications. Effects of Interruptions Detrimental Effects of Interruptions Introduction Interruption has been an active area in human-computer interaction research for some time. A comprehensive review was provided by McFarlane and Latorella (2002). Interruptions are typically defined as a change or disturbance in a process or in people’s activities.(Cooper & Franks, 1993; McFarlane & Latorella, 2002) Interruptions are categorized along different dimensions by different researchers, such as source, effect, content, applicability, and duration by Cooper & Franks (1993) and individual properties, methods, meaning, source, channel, change, and effect by McFarlane and Latorella (2002). Significant research has been expelled in determining how to classify, prevent, minimize, and provide tools to help users deal with interruptions. However, there is little understanding how interruptions can be exploited for The effects of interruptions are generally described as negative Users perceive an interrupted task as being more difficult to complete than an uninterrupted task (Bailey, Konstan, & Carlis, 2000). An interruption is also thought to take longer to process and return back to task when it is unrelated to the task at hand (Cutrell, Czerwinski, & Horvitz, 2001). The added memory load seems to make it difficult for a task to be resumed. It also becomes difficult to remember what task was being processed before the interruption. (Burmistrov & Leonova, 1996; Dix, Ramduny, & Wilkinson, 1995). Further, the complexity of the task being interrupted effects the disruptiveness of an interruption. Interrupting complex tasks inhibits performance, and has no effect on simpler tasks (Burmistrov & Leonova, 1996). Interestingly, people can recall details about interrupted tasks better than uninterrupted tasks.(McFarlane & Latorella, 2002) People also have individual differences in their ability to respond and manage interruptions (McFarlane & Latorella, 2002). Interruptions also affect performance. Users are thought in general to perform slower on interrupted tasks (Bailey et al., 2000), although some evidence exist that an interruption may actually speed up task completion (Zijlstra,