Criterion-Referenced Setting of Cut-Points and Evaluation of Fitness

Most of the limitations listed above are overcome with a criterion-referenced evaluation approach, whereby a test taker’s performance is compared with an absolute criterion related to whether a child meets a minimal necessary physical fitness level. In contrast with the norm-referenced approach, because the criterion is defined independently, it is not impacted by changes in a population that occur over time or in the level of fitness of a specific population. Limitations related to genetic differences or the potential for discouraging unfit participants also can be minimized with this approach.

Many methods have been developed for setting performance standards (Cizek, 2001; Livingston and Zieky, 1982), but for criterion-referenced evaluation of health-related fitness, the health outcome–centered method (Zhu et al., 2011) has predominated. Basically, this method involves linking health-related fitness performance with a particular (set of) health outcome measure(s). Specifically, this approach identifies a level of test performance that discriminates, with acceptable specificity and sensitivity, between participants who have and do not have a defined health characteristic. An example is identifying a level of performance on a measure of cardiorespiratory endurance that discriminates between groups of youth who have or do not have an at-risk score for metabolic syndrome.

Steps completed before developing cut-points for the health-outcome centered method include determining the components of health-related fitness (e.g., cardiorespiratory endurance) and selecting valid, reliable criterion measures and field tests and health outcomes or markers. Field tests are selected because even though criterion measures (criterion-referenced standards) are the most accurate measure of a construct, they often are more expensive and time-consuming and require sophisticated equipment. Field tests are more practical, less costly, and less time-consuming for mass testing. It is important, however, to determine the validity and reliability of field tests by deriving the predictive relationship and determining its consistency with the selected criterion measure.

The selection of health outcomes can be based on the expected relationship between field tests and health markers or outcomes. Health is a construct, so there are many possible health outcome measures, such as mortality, a single risk factor (e.g., blood pressure), or a group of risk factors (e.g., metabolic syndrome). Because no specific measure is considered superior, it is advisable to use multiple outcome measures to validate the test results when possible. Those selecting cut-points also need to identify health outcome measures using existing standards (e.g., a systolic blood pressure level of ≥103 mmHg for a 5-year-old child 104 cm in height) and make adjustments for specific populations if needed.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement