Data and Code for: Identifying Students at Risk Using a New Math Skills Assessment

Math skills are critical for success in economics courses. Introductory courses require students to do unit conversions, understand percentages, and translate word problems into mathematics. Students must also be able to work with graphs and solve algebraic equations. Intermediate level courses in theory and econometrics often require students to differentiate functions, solve optimization problems, and compute integrals. Students without these skills usually perform poorly or drop their economics courses midway through the semester.<br><br>If students-at-risk can be identified at the beginning of the term, they can be given the support they need to succeed. This support may take the form of a nudge to attend office hours regularly or use a department tutoring service. It may also involve enrolling them in a supplemental course during the term or even postponing the economics course until they take an alternative dedicated to significantly improving their math skills (Foster et al., CTREE 2019). All students therefore receive the support they need from the very beginning of the semester, and faculty can dedicate their scarce classroom time to teaching economics instead of basic math. <br><br>In this paper we present two new assessments that economics instructors can use to evaluate students and identify those with weak math skills early in the semester. The MESA-Foundations (Mathematics for Economics Skills Assessment - Foundations) instrument tests explicitly defined learning goals in arithmetic, graphs, and algebra. MESA-Intermediate shares the MESA-Foundations questions on algebra and graphs and adds a set of questions that evaluate calculus skills. Both assessments are multiple choice and are designed to be completed in under 45 minutes without a calculator. In our paper we show how our assessments, given at the beginning of the term, predicted performance of 900 students in an introductory microeconomics class and 300 students in an intermediate microeconomic theory class. We analyze course completion and exam grades in both courses. We use Lasso estimation to identify optimal thresholds for interventions described above.