MouseTrack: A Web-Based Usability Tool
Ernesto Arroyo and Ted Selker
MouseTrack is a web logging system that tracks mouse movements on websites. It includes a visualization tool that displays the mouse cursor path followed by website visitors. MouseTrack has the capability to dynamically augment web pages based on detected browsing behaviors. It helps web site administrators run usability tests and analyze the collected data. The demo includes a design case that describes how the tool can be used for webpage personalization.
Webpage: http://ctxt.media.mit.edu/php5usertrack/
Studies in Disruptive Interruptions
Ernesto Arroyo and Ted Selker
Disruption is an important issue in the design of self-adaptive interfaces. This project attempts to make a computer intelligent enough to select the appropriate sensorial modality. This project investigates which modality is the most efficient and at the same time, the least disruptive. Two interruption modalities are studied: heat and light. This research adds to previous research by showing that there is an effect on performance caused by interruption modalities: for example, thermal modality produced a larger decrease in performance than visual modality; this modality has a greater disruptive effect on interrupted tasks than light. Disruptiveness and performance measures agree that heat causes more of a detrimental effect than light when used as an interruption.
Interruption Manager
Shawn Sullivan and Ted Selker
The Interruption Manager is a system designed to allow both users and experimenters to easily control the manner in which programs interact. Combining several different insights into dealing with the disruptiveness of interruptions, the Interruption Manager provides many different combinations of interruption presentation and control. In addition, the Interruption Manager bases its actions on what programs are open on the computer, what windows are open and active, and the user’s keyboard and mouse usage, making it application independent. Interruption Manager Webpage
Predicting Web Interruptibility
Ted Selker and Ernesto Arroyo
People use cues for determining appropriate moments for interruption. If somebody just finished taking on the telephone, that action might be a good indication to knock at their door. What cue detail level is necessary for people to make appropriate interruption decisions? We explored if humans are capable of inferring interruption times with an experiment that presented a series of screenshots from people navigating Websites. Subjects were asked to select three possible interruption times based on mouse activity. Mouse movement trajectories can indicate how information is interpreted while browsing a Webpage. Slow and arched trajectories as users move their mouse would indicate an ambiguous state of mind.
EngagementTracker
Engagement Tracker is a system that measures users engagement while browsing web pages using three main components. Web page structure analysis, Mouse movement analysis, and Video synchronization.
Cellular Interruption
This application provides a way to design experiments that study the disruptive effects of interruption. The application is based on a cell phone MIDlet that periodically interrupts users with incoming messages intended to divert their attention from an ongoing activity. The interruptions to the users are controlled by an experimenter who may choose to interrupt the users during specific activities. The application logs data about users’ context as well as information about the users’ responses to the messages (including the level of disruption caused by the interruption). The application may be used to conduct a wide variety of experiments to study human response to interruption. The results of a pilot study using the application, show that random interruptions, that are unrelated to users’ goals and activities, are significantly more disruptive than related interruptions.