Panel Paper: An Experimental Evaluation of a Computer-Assisted, Modular Approach to Developmental Math

Friday, November 3, 2017
Haymarket (Hyatt Regency Chicago)

*Names in bold indicate Presenter

Michael Weiss and Camielle Headlam, MDRC

The Problem:

Community colleges play a vital role in postsecondary education – they enroll nearly half of all college students and, because of their open admissions policies and low cost, are accessible to millions of adults who might otherwise lack the preparation or financial means to pursue higher education. Unfortunately, this open access does not always translate into academic success. Nationally, three year graduation rates are around 20 percent. Rates are especially low for students who enter with developmental (remedial) course needs.

The Program:

Community colleges are looking for solutions to help students complete developmental (remedial) math — a course sequence viewed my many as the single biggest barrier to graduation. Some are offering computer-assisted, modular developmental math courses that allow students to earn credits incrementally and move through the curriculum at their own pace.

One of these modularized courses, ModMath, was created at Tarrant County College (TCC) in Texas. It reorganizes the content of TCC’s two semester developmental math course sequence into a set of six modules, each of which is five weeks long. The four primary components of the ModMath intervention are: (1) a diagnostic assessment that places students in a starting module; (2) individual registration into modules; (3) computer-based (adaptive) instruction delivered through an instructional software program; and (4) personalized, on-demand assistance in class from an instructor and class aide.

Research Design:

This study is a randomized field trial with a sample of over 1,400 students requiring developmental math. Participating students were randomly assigned to be a part of the program and have the opportunity to participate in ModMath or to experience the college’s regular math course (typically a lecture-based course).

Findings After One Semester:

As shown in Table 1, on the positive side, on average, program group students were 10 percentage points closer to completing the developmental math sequence than control group students — 25 percent of the way through it compared with 15 percent, a finding largely driven by the fact that students in ModMath are able to complete 1 credit modules, where students in the control group either completed an entire 3 credit course or earned zero credits. This advantage did not translate into other measures of progress. For example, program group students were not more likely than the control group to pass the halfway mark in the developmental math sequence. More than 70 percent of students in the study, in either group, were unable to pass this benchmark in the first semester. Moreover, ModMath had a very small but statistically significant and negative impact on the percentage of students who completed the entire developmental math sequence during their first semester, 0.4 percent for the program group compared with 1.9 percent for the control group.

What’s To Come:

By the time of the November APPAM conference, findings for the full sample will be available through at least 1.5 years – a more reasonable timeframe for students to have completed the entire developmental course sequence.