| Friday, February 16 | |
| PS2 Poster Session 2 and Refreshments |
Fri, Feb 16, 5:15 PM - 6:30 PM
Salons F-I |
Curating and Visualizing Big Data from Wearable Activity Trackers (303691)Luke J. Burchill, OHSU Knight Cardiovascular InstituteNora Fino, OHSU-PSU School of Public Health Jodi Lapidus, OHSU/PSU School of Public Health Jessica Minnier, Oregon Health & Science University *Meike Niederhausen, OHSU-PSU School of Public Health Yuliang Wang, University of Washington Keywords: big data, visualizations, R, tidyverse, wearable devices Fitness trackers have become a popular device for researching associations between physical activity and health status. We will share challenges in curating and analyzing data from wearable devices, as well as activity data visualizations and the R code used to create them. The data are from a recent 5-month workplace study with 431 healthy volunteers that collected minute-by-minute activity from a wearable device along with biometric and cardiometabolic measures at the start and end of study. Tracker data included heart rate, steps, calories, body states (sleep, inactive, light activity, moderate activity, and off wrist), workouts (walk, run, and cycle), and sleep metrics. Visualizations played a key role in cleaning the data, detecting outliers, and identifying issues in how the data had been recorded. They were also essential in helping collaborators understand the complex relationships between the many variables and determine which were the most relevant and reliable for analyses. This poster will be of interest to researchers analyzing data tracking activity over time, as well as R users who want to learn more about wrangling and visualizing data with tidyverse and ggplot.
|
|
