Research
Our work starts from measurement methodology and extends to population health and sports data analytics. Every project asks the same question: how far can we trust this measurement?
We compare steps, energy expenditure, and sleep from consumer activity trackers, smartwatches, and smartphones against criterion measures, and have contributed to standards for analyzing and reporting validation studies.
Using national datasets such as KNHANES, the Korean Genome and Epidemiology Study (KoGES), the NHIS Senior Cohort, and the Korean Children and Youth Panel Survey, we study how physical activity relates to chronic disease, cognitive function, fractures, and mortality.
We treat physical activity, sedentary time, and sleep as parts of one day, derive activity-rhythm and sleep-timing indicators from raw accelerometer data, and relate them to healthy aging.
We analyze professional baseball sabermetrics, tennis match records, and sports-interest data from a measurement perspective, checking data quality and metric reliability first.
We develop cross-validated prediction equations for maximal heart rate, VO₂max, and energy expenditure, and design field tests such as the physical fitness test for Korean police officers.
We use G-theory to estimate how many days or weeks of monitoring are needed for stable estimates, applied to adolescent Fitbit data, athlete wearable data, and match-charting data.
