Predictive Validity

Predictive validity refers to the degree to which scores on a test or assessment are related to performance on a criterion or gold standard assessment that is administered at some point in the future. Predictive validity is often considered in conjunction with concurrent validity in establishing the criterion-based validity of a test or measure. Although concurrent validity refers to the association between a measure and a criterion assessment when both were collected at the same time, predictive validity is concerned with the prediction of subsequent performance or outcomes. Educators, researchers, and practitioners are often interested in how well a test or assessment will forecast an individual’s future performance in a particular domain; therefore, predictive validity is an important aspect for demonstrating the technical adequacy and practical utility of a test or measure.

Predictive validity is typically established using correlational analyses, in which a correlation coefficient between the test of interest and the criterion assessment serves as an index measure. Multiple regression or path analyses can also be [Page 1289]used to inform predictive validity. Because the administration of the test and the criterion assessment may be separated by several weeks, months, or years, it should be noted that predictive validity coefficients are often weaker than concurrent validity coefficients due to maturation, learning, or other variables associated with the passage of time between the assessment occasions.

Several examples of applications of predictive validity to education research and practice can be considered. A measure of toddlers’ receptive vocabulary might be evaluated for its ability to predict scores on an assessment of vocabulary proficiency at school entry. An assessment designed to evaluate kindergarten students’ emergent literacy skills might be evaluated in terms of how well it predicts their performance on a reading assessment in a subsequent grade. A test of math skills administered at the beginning of the school year might be evaluated for its ability to predict students’ scores on a test of overall mathematics proficiency taken later in the year. Or, an assessment of high school students’ study habits might be evaluated for the strength at which it predicts rates of school completion or performance on college entrance examinations.

Because predictive validity is a key part of demonstrating the technical adequacy of a test or measure, it is important for test developers to report results of predictive validity studies. Additionally, researchers or educators who are considering the use of a particular test or measure for predicting future performance should determine whether the test has demonstrated evidence of predicting outcomes or performance in the domain of interest. This information is often found in technical support materials provided by the publisher or in independent research studies that have evaluated the predictive validity of the test.

See also Concurrent Validity; Correlation; Criterion-Based Validity Evidence; Tests; Validity; Validity Coefficients