[Mobilizeplans-starstudents] Mobilize Center Seminar, Eric J. Daza, Nov 2

[Diane Bush]

Our next Mobilize Center Seminar is scheduled for Thursday, November 2nd, and features Eric J. Daza, postdoctoral fellow from the Stanford Prevention Research Center. He will be presenting “Causal analysis of self-tracked health data using an n-of-1 counterfactual framework”.

The Mobilize Center seminars are held once a month.  Please check Mobilize Events<http://mobilize.stanford.edu/events/> for updates.

We look forward to seeing you in November and at future seminars!

TITLE:
Causal analysis of self-tracked health data using an n-of-1 counterfactual framework

WHEN:
Thursday, November 2
noon - 1 pm

WHERE:
Y2E2 300, Stanford University

Abstract:
Mobile apps and wearable devices are empowering more and more individuals to investigate their own health (e.g., chronic illnesses/conditions, fitness/diet/nutrition goals, health-related behaviors). N-of-1 randomized trials (N1RTs) are single-subject crossover trials that many “self-trackers” can use to assess the effects of possible causes, triggers, or risk factors discovered using such observational data. N1RT data are structured time series: A treatment is administered repeatedly within a treatment period, outcomes may be autocorrelated or express a time trend, and the treatment effect may carry over into subsequent periods, or take time to stabilize/decay upon administration/removal. All of these characteristics complicate causal modeling such that naïve causal discovery using observational time series data may not properly define the types of effects to expect from an N1RT. In this talk, I will formalize N1RT design in terms of potential outcomes (i.e., counterfactuals) in order to conduct exploratory n-of-1 observational studies (N1OSs; i.e., without randomized treatments or predefined treatment periods) for generating causal hypotheses that better target N1RT effects. I will then use this framework to see if changes in physical activity may have affected my average body weight over a six-year timespan, discuss N1OS analytical challenges, propose a basic N1OS approach, and introduce a forthcoming N1OS on causal prediction of migraine headaches.

To keep up-to-date on upcoming speakers and the dates, visit Mobilize Events<http://mobilize.stanford.edu/events/>.


Diane Bush
Assistant to Professor Scott Delp
NMBL, Mobilize Center, OpenSim
Stanford University
dbush1 at stanford.edu<mailto:dbush1 at stanford.edu>


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://simtk.org/pipermail/mobilizeplans-starstudents/attachments/20171026/732c38f7/attachment.html>