![]() The framework identifies six steps of the physical activity assessment process among which efforts might be made to reduce measurement errors with self-report. 2) identifying information that questionnaire users should know and/or should ask to reduce measurement error and 3) providing suggestions to enhance the use of questionnaire to assess physical activity behavior in research and practice settings‥Ī conceptual framework for reducing errors with self-report assessment of physical activity is presented to summarize the workgroup discussions ( Figure 1). The workgroup discussions summarized in this paper focused on: 1) recommendations for researchers, practitioners, and organizations to minimize errors with self-report questionnaires. This effort can be accomplished across many steps in the physical activity assessment process and will involve a variety of individuals performing different roles across the steps. To improve instrument accuracy in this context, it is clear that one should identify and combat the underlying sources of measurement error and/or correct for it when it cannot be averted. Investigators try to reduce systematic error by assuring instruments are valid and reliable in the study population measured. Systematic errors may have equally damaging effects on the study as random errors and can cause the measured mean scores to appear higher or lower than the true scores. Systematic error may be caused by reporting biases or a mismatch between the scope of a questionnaire relative to a reference measure chosen to estimate the true score. Systematic error, on the other hand, is caused by factors that uniformly change the mean value of a score across a study sample. Investigators try to reduce random error by standardizing the testing conditions, by enhancing study design (i.e., incorporating replicate measures), and by assuring a sufficient number of participants are included in a study to minimize the variability of scores around the mean. While random error generally does not affect the mean value for a group, it is unpredictable in how it can affect the variation in scores. ![]() They are generally attributable to unreliable reporting and natural variation in behaviors over time. Random error exists in all questionnaires. Two sources of error are common in a self-report instrument: random error and systematic error 1. ![]() Thus, it is important to understand the types and sources of measurement error in self-report instruments and to identify ways to reduce and/or correct for measure error. If severe enough, the attenuation may limit the ability of a study to detect actual relationships between the variables. Measurement error can lead to biased estimates of a behavior and may attenuate or overestimate relationships between a physical activity score with other variables. All instruments, including self-report questionnaires, have some level of measurement error that, if unchecked, can lead to uncertainty of measures and flawed inference.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |