Psychometric report on the Knowledge for Teaching Elementary Fractions test administered to elementary educators in six states in fall 2016

This report provides evidence of the substantive and structural validity of the Knowledge for Teaching Elementary Fractions Test. Field-test data were gathered with a sample of 277 elementary educators, including teachers, administrators, and instructional support personnel, in fall 2016, as part of a larger study involving a multisite cluster-randomized trial evaluation design to investigate the effects of lesson study and a fractions resource toolkit on classroom instruction and student achievement in fractions. The web-based test was designed to measure mathematical knowledge for teaching (MKT) in the domain of fractions at the elementary level. we found that the Knowledge for Teaching Elementary Fractions test measures a dominant factor, supporting unidimensionality in the data. The purpose, or intended use, of the Knowledge for Teaching Elementary Fractions test is to produce ability estimates that can be used to investigate baseline equivalence of groups of educators in four treatment conditions, to serve as a covariate in models estimating the effect of the intervention on MKT, as well as to investigate MKT as a potential moderator of the effect of the program on teachers and students. Reliability, test-information, and item-discrimination estimates appear to fit the intended purpose of the test, although further validation will be necessary to determine whether the test is well suited for its intended use.