BACKGROUND: Physical activity recall instruments provide an inexpensive method of collecting physical activity patterns on a sample of individuals, but they are subject to systematic and random measurement error. Statistical models can be used to estimate measurement error in activity recalls and provide more accurate estimates of usual activity parameters for a population. METHODS: We develop a measurement error model for a short-term activity recall that describes the relationship between the recall and an individual's usual activity over a long period of time. The model includes terms for systematic and random measurement errors. To estimate model parameters, the design should include replicate observations of a concurrent activity recall and an objective monitor measurement on a subsample of respondents. RESULTS: We illustrate the approach with preliminary data from the Iowa Physical Activity Measurement Study. In this dataset, recalls tend to overestimate actual activity, and measurement errors greatly increase the variance of recalls relative to the person-to-person variation in usual activity. Statistical adjustments are used to remove bias and extraneous variation in estimating the usual activity distribution. CONCLUSIONS: Modeling measurement error in recall data can be used to provide more accurate estimates of long-term activity behavior.
BACKGROUND: Physical activity recall instruments provide an inexpensive method of collecting physical activity patterns on a sample of individuals, but they are subject to systematic and random measurement error. Statistical models can be used to estimate measurement error in activity recalls and provide more accurate estimates of usual activity parameters for a population. METHODS: We develop a measurement error model for a short-term activity recall that describes the relationship between the recall and an individual's usual activity over a long period of time. The model includes terms for systematic and random measurement errors. To estimate model parameters, the design should include replicate observations of a concurrent activity recall and an objective monitor measurement on a subsample of respondents. RESULTS: We illustrate the approach with preliminary data from the Iowa Physical Activity Measurement Study. In this dataset, recalls tend to overestimate actual activity, and measurement errors greatly increase the variance of recalls relative to the person-to-person variation in usual activity. Statistical adjustments are used to remove bias and extraneous variation in estimating the usual activity distribution. CONCLUSIONS: Modeling measurement error in recall data can be used to provide more accurate estimates of long-term activity behavior.
Authors: Gregory J Welk; Youngwon Kim; Bryan Stanfill; David A Osthus; Miguel A Calabro; Sarah M Nusser; Alicia Carriquiry Journal: Med Sci Sports Exerc Date: 2014-10 Impact factor: 5.411
Authors: Pedro F Saint-Maurice; Youngwon Kim; Paul Hibbing; April Y Oh; Frank M Perna; Gregory J Welk Journal: Am J Prev Med Date: 2017-06 Impact factor: 5.043
Authors: Ruth H Keogh; Pamela A Shaw; Paul Gustafson; Raymond J Carroll; Veronika Deffner; Kevin W Dodd; Helmut Küchenhoff; Janet A Tooze; Michael P Wallace; Victor Kipnis; Laurence S Freedman Journal: Stat Med Date: 2020-04-03 Impact factor: 2.373
Authors: Natalie Anderton; Megan E Newhouse; Barbara E Ainsworth; Ingrid E Nygaard; Marlene J Egger; Janet M Shaw Journal: J Phys Act Health Date: 2014-05-06
Authors: Joshua N Sampson; Charles E Matthews; Laurence Freedman; Raymond J Carroll; Victor Kipnis Journal: J Appl Stat Date: 2016-03-17 Impact factor: 1.404
Authors: Charles E Matthews; Sarah Kozey Keadle; Joshua Sampson; Kate Lyden; Heather R Bowles; Stephen C Moore; Amanda Libertine; Patty S Freedson; Jay H Fowke Journal: Med Sci Sports Exerc Date: 2013-08 Impact factor: 5.411