An estimate of random errors of measurement related to the use of parallel, equivalent forms of an assessment tool (e.g., Form A versus Form B).