Abstract
Mobile technologies create new opportunities to develop personalized human-computer interfaces that respond to everyday changes in behaviors. Such interventions are fueled by computational models that need information on behaviors in the real-world. Ideally, sensors embedded on mobile devices could measure these behaviors. But we currently need self-report to capture subjective experiences that sensors cannot measure directly (e.g., fatigue, pain, and productivity). Ecological momentary assessment (EMA) or experience sampling method (ESM) is one such approach that enables in-situ self-report data collection using smartphones. In EMA, users are prompted several times a day on their phones to answer sets of multiple-choice questions. The in-situ nature of EMA reduces recall bias compared to long-form traditional questionnaires. Additionally, the repeated nature of EMA helps capture variations in experiences unique to each user. But, this method of data collection poses a heavy burden on the end users, impacting both the quality and quantity of collected survey data. Aditya’s work is driven by a fundamental question: Given this trade-off between user burden and data quantity, how can we scale experience sampling in real-world settings?
In this talk, Aditya will present a novel (and first-of-its-kind) EMA approach called μEMA that uses micro-interactions on the Smartwatch to collect survey data in natural settings – both at high-frequency and at large-scale. The μEMA restricts EMA interruptions to single, cognitively simple questions that can be answered on a smartwatch with a single tap – a quick, glanceable microinteraction like checking time. Because of this microinteraction, μEMA permits substantially higher interruption than EMA without as much user burden. Aditya will present results from several pilot studies and an year long longitudinal evaluation of μEMA to show that this method: 1) yields higher response rates from end users, 2) poses less burden on users, and 3) produces accurate information on user’s momentary experiences. Finally, Aditya will discuss potential applications of μEMA method for data collection, user behavior modeling, and personalization.
About Aditya
Aditya is an interdisciplinary researcher interested in human-computer interaction and behavior science, with a focus on measuring and modeling user experiences. He is currently working as a Sr. Researcher at MongoDB focusing on developer experiences and growth. Previously he was a Research Scientist at Spotify (as a member of Human-AI Interaction Lab), where he developed and studied interactive recommender systems and computational approaches to help smaller podcasters and newer music content grow audience on the platform. He completed his PhD in Computer and Health Sciences from Northeastern University, Boston, MA, where his research led to the development of novel experience
sampling tools, crowdsourcing platforms, and open source tools to perform multi-level modeling and data annotation on high-frequency longitudinal mobile data. His research has been published in peer-reviewed competitive venues including ACM IMWUT (UbiComp), ACM CSCW, ACM IUI, ACM WebSci, ACM CHIPLAY, IEEE PerComm, NeurIPS Workshops, Behavior Research Methods, Journal of Medical Internet Research, and Translational Behavior Medicine and has won two best paper awards.
This Event will be held in person at 11 Leon Street, Boston, MA. Ryder Hall, Room 180 (first floor), and remotely.