Association for Public Policy Analysis & Management

Our paper evaluates the recent adoption of a technology-centered curriculum that has replaced traditional 12^th grade math courses in over 140 high schools across the state of Tennessee.  The SAILS (Seamless Alignment and Integrated Learning Support) program allows students who, based on their 11^th grade ACT math sub-score, would otherwise be placed into developmental math once in college, to instead take the identical course during their senior year of high school.  Students passing the SAILS course in 12^thgrade are then exempt from math remediation, and can begin college in credit-bearing math courses if they attend any in-state public college or university. This research study examines the effects of enrolling in a SAILS math class on students from different backgrounds, and with differing levels of academic preparation.

Using data provided by the Tennessee Higher Education Commission (THEC) and the Tennessee Department of Education, we explore the effects of enrolling in SAILS for two cohorts of high school seniors (2012 and 2013).  We use a regression discontinuity design that compares outcomes for students whose ACT scores are just above and below the cutoff score of 19 for placement into SAILS. This allows us to compare students assigned to SAILS to their peers scoring one or two points higher on the ACT math test who did not participate in SAILS, and to compare SAILS students to their peers also with ACT scores below a 19 attending a high school that does not offer SAILS.  Specifically we ask: does enrolling in SAILS during the senior year of high school result in higher rates of high school graduation, college attendance, and eventual success in college-level math, as compared to similar students not assigned to SAILS? Do these results differ by student sub-group (prior level of academic preparation, age, SES, etc.)?  We also explore which of the five SAILS math competencies are particularly difficult for students to pass, and whether differences exist by subgroup.  The results from the first cohort indicate varied levels of effectiveness among different subgroups of students enrolled at different types of high schools (small, large; urban, rural) throughout the state, providing insight into the extent to which the particular instruction and delivery methods of developmental math courses affect subsequent student academic outcomes.  The results of this study are particularly timely, as the program aims toward scaling to reach all high school students with ACT math exam scores below a 19 (~30,000) by the 2015-16 academic year. We discuss potential implications for this scale-up in light of our findings.