Standardized formative mathematics assessments typically fail to capture the depth of current
standards and curricula. Consequently, these assessments demonstrate limited utility for
informing the instructional implementation choices of teachers. This problem is particularly
salient as it relates to the mathematical problem solving process. The purpose of this study was
to develop and evaluate the psychometric characteristics of Structural-Symbolic Translation
Fluency, a curriculum-based measure (CBM) of mathematical problem solving. The
development of the assessment was based on previous research describing the cognitive process
of translation (Mayer, 2002) as well as mathematical concept development at the quantitative,
structural, and symbolic levels (Dehaene, 2011; Faulkner, 2009; Griffin, 2004).
Data on the Structural-Symbolic Translation Fluency assessment were collected from 11
mathematics and psychometrics experts and 42 second grade students during the spring of 2016.
Data were analyzed through descriptive statistics, frequencies, Spearman-Brown correlation,
joint probability of agreement, Pearson correlation, and hierarchical multiple regression.
Psychometric features of interest included internal consistency, inter-rater reliability, test-retest
reliability, content validity, and criterion-related validity. Testing of the 9 research questions
revealed 9 significant findings. Despite significant statistical findings, several coefficients did
not meet pre-established criteria required for validation. Hypothesized modifications to improve
the psychometric characteristics are suggested as the focus of future research