By Anind K. Dey, Carnegie Mellon University, Embedded Assessment Principal Investigator
One of the challenges we are facing on our project is the attempt to conduct research across two different disciplines or domains and to demonstrate credible accomplishments in both. On one hand, there is the world of technical achievement in computer science, which myself and the members of my team are very comfortable with. We know what an appropriate contribution is in this field and what entails appropriate demonstrations of such contributions. On the other hand, is the focus on clinical outcomes, and is an area where we feel much less comfortable.
From our interactions with grantees from Round 1 of ProjectHealthDesign, it's clear that others share the same discomfort, although they may feel more comfortable on the clinical side and less so on the technical or user interface side.
Our goal for our project is to better understand how embedded sensing, a technical construct, can be used to help elders and related stakeholders reflect on sensed aspects of everyday behaviors to determine whether there is a change in an elder's cognitive health. We have chosen this goal specifically, as it matches our understanding of a technical contribution in our field of human-computer interaction.
In addition, a clinical goal which we will not achieve in our project would be to validate that the measures obtained from embedded sensing are as accurate, if not more accurate, than existing assessment techniques. While we would love to work toward this goal, it seems beyond our abilities with respect to the numbers of people we can work with, and the timeline of ProjectHealthDesign.
I would love to hear from my clinical colleagues and others about methods or approaches that would allow for both technical and clinical contributions. In this age of increased interest in applying technology to clinical domains, a better understanding of how to support both types of work simultaneously would enable many of us struggling on both sides of this problem. What are some approaches we should pursue?
Comments