CAMBRIDGE, MA—Computer users realize that running too many apps at the same time can make their computer slow down. The same happens in the human brain. Too many apps can lead to cognitive overload—a situation that can drag down employee productivity. The problem may be worse than imagined. According to a recent survey, seven in 10 computer users spend an average of 32 days a year trying to navigate between the maze of apps and services at their disposal.
Krysta Chauncey, an applied cognition scientist at Draper, believes that tailoring the interface for an individual user’s needs can address choice overload. “Overly complex software applications can slow you down and distract you, making you less productive—the exact opposite of what they’re supposed to do,” Chauncey said.
To address this challenge, Chauncey and a team of Draper engineers set out to understand how people and machines adapt to each other. They tasked a group of computer users to work with a data analytics software application and then applied principles of human-computer interaction, psychological theory and graphical user interface design to assess how the users did.
In the test, participants were given 10 minutes to engage in an open-ended data analysis task with specific interface adaptations. During the task, participants explored the interface and verbalized any action or change they would like to perform. The dashboard was updated with the necessary changes.
Researchers found that users with a high tendency to engage in and enjoy effortful cognitive tasks—also known as ‘need for cognition’ (NFC)—sought out more data and were drawn toward exploring deeper into specific data sets. Conversely, low NFC users tended to be drawn toward items that were visually interesting and quickly avoided any elements they felt they did not understand.
“Some people want lots of data, while others prefer prompts, widgets, tips and other aids to data analysis,” Chauncey said. “The goal in advancing UX is to move toward a co-adaptive relationship where both the human and the software agent adjust behavior and information presentation based on context, state, trait and task.”
Customizing software applications according to user behavior isn’t new, but Draper’s research takes a deeper look at how a machine could adapt to individuals, according to a paper published by the researchers. “While many applications contain some form of customization, such as recommendations from Netflix, product suggestions from Amazon or word suggestions in Google, personalizing more effectively beyond recommendation content requires a more specific and concrete understanding of each audience member for particular applications,” they said.
The implementation of this ability to adapt in real time is called the Draper Continuously Tailored Software.
To Chauncey, adapting to users in real time can combine the power of high-featured software with the accessibility of simple software. “We set out to measure how a machine adapting to individuals would affect their performance, and we found that we could mitigate cognitive overload and supplement natural processing capabilities.”
This work in human-centered computing is part of Draper’s growing Human-Centered Solutions portfolio. The portfolio includes a “take me home” button designed to return astronauts safely to an orbiting space station; a wearable technology called isaWear (Immersive Situational Awareness) that helps the wearer recognize more data from their surroundings and understand them faster; and a sensor suite called Drowsy Driver Detection that can be built into a car driver’s headrest for detecting brain waves that signal the beginnings of drowsiness and trigger an alert to the driver.