Multiple Measures Placement Using Data Analytics: An Implementation and Early Impacts Report
By Elisabeth A. Barnett, Peter Bergman, Elizabeth Kopko, Vikash Reddy, Clive R. Belfield, and Susha Roy, with Dan Cullinan | September 2018
Because institutions often rely solely on standardized placement tests to determine students’ college readiness, many incoming community college students who could have succeeded in entry-level courses are required to take remedial math or English first. Referring these students to developmental education needlessly stalls their progress toward a degree, as they are forced to sink time and money into classes that do not earn them college credit.
CAPR is studying whether combining multiple measures, including placement test results and high school GPA, into a data analytics algorithm allows colleges to more accurately predict students’ performance in college-level math and English and thus place them in the courses that will best support their progress toward a degree. This report describes early results from CAPR’s experimental study of 13,000 students at seven New York community colleges who were randomly assigned to be placed using either standardized placement tests alone (control group) or multiple measures (program group).
While implementing the alternative placement system was more complex than expected, all seven colleges were successful in implementing it. And early impacts results indicate students placed using multiple methods were more likely to place into and complete college-level courses in their first term.
This is the first of two reports. The final report, to be released in 2019, will include data on students’ completion of introductory college-level courses, persistence, and accumulation of college credits over the longer term.