Examining Construct Validity of the Quantitative Literacy VALUE Rubric in College-Level STEM Assignments.

John P. Connolly, Ph.D. The University of Texas at Arlington Abstract Data-driven decision making is increasingly viewed as essential in a globally competitive society. Initiatives to augment standardized testing with performance-based assessment have increased as educators progressively respond to mandates for authentic measurement of student attainment. To meet this challenge, multidisciplinary rubrics were developed as a method of scoring student work samples. The current study utilized confirmatory factor analysis to examine ratings of student work (N = 245) using the Quantitative Literacy VALUE Rubric from the Association of American Colleges and Universities. The study examined a conceptual model of the six skill measures from the rubric to validate whether, taken together, they are reliable measures of a single general construct—Empirical and Quantitative Skill (EQS), a Texas Core Curriculum objective. The model confirmed that the six measures in the rubric (Interpretation, Representation, Calculation, Application/Analysis, Assumptions, and Communication) appeared to describe a single construct. Results support using the Quantitative Literacy VALUE Rubric for assessing EQS.