Author:
Kim Steven,Essert Christopher
Abstract
An accurate and reliable measurement is important in exercise science. The measurement tends to be less reliable when
subjects are not professional athletes or are unfamiliar with a given task. These subjects need familiarization trials, but
determination of the number of familiarization trials is challenging because it may be individual-specific and task-specific.
Some participants may be eliminated because their results deviate from arbitrary ad hoc rules. We treat these challenges
as a statistical problem, and we propose model-averaging to measure a subject’s familiarized performance without fixing
the number of familiarization trials in advance. The method of model-averaging accounts for the uncertainty associated
with the number of familiarization trials that a subject needs. Simulations show that model-averaging is useful when the
familiarization phase is long or when the familiarization occurs at a fast rate relative to the amount of noise in the data.
An applet is provided on the internet with a very brief User’s Guide included in the appendix to this article.
Keywords: Familiarization; reliability; accuracy; model-averaging; Akaike Information Criterion
Reference20 articles.
1. Akaike, H. (1974). A new look at the statistical model identification. IEEE Transactions on Automatic Control, 19(6), 716-723.
2. Bailer, A. J., Noble, R. B., & Wheeler, M. W. (2005). Model uncertainty and risk estimation for experimentalstudies of quantal responses. Risk Analysis, 25(2), 291-299.
3. Beckham, G., Lish, S., Disney, C., Keebler, L., DeBeliso, M., & Adams, K. J. (2019). The reliability of the seatedmedicine ball throw as assessed with accelerometer instrumentation. Journal of Physical Activity Research, 4(2),108-113.
4. Buckland, S. T., Burnham, K. P., & Augustin, N. H. (1997). Model selection: An integral part of inference. Biometrics, 53, 603-618.
5. Burnham, K. P., & Anderson, D. R. (2002). Model selection and multimodel inference: A practical information-theoretic approach. New York: Springer-Verlag.