Community colleges now serve over one-third of our nation’s postsecondary students each year. Because they have open admissions and are relatively low cost, they enroll larger percentages of low-income students than four-year institutions. Unfortunately, as enrollment in these colleges has increased, students’ success rates have not kept pace. One of the major barriers for academically underprepared students is the need to pass developmental (or remedial) math classes. These classes do not offer college credits, and rates of completing and passing them are low. Learning communities are a popular strategy for moving students through the developmental math sequence. They enroll a cohort of students in two classes and often incorporate shared assignments and curricula, collaboration between faculty teaching pairs, and connections to student support services.

Queensborough Community College and Houston Community College are two large, urban institutions that offer learning communities for their developmental math students, with the goals of accelerating students’ progress through the math sequence and of helping them to perform better in college and ultimately earn degrees or certificates. They are two of six colleges participating in the National Center for Postsecondary Research’s Learning Communities Demonstration, in which random assignment evaluations are being used to determine the effects of learning communities. At Queensborough, classes in all levels of developmental math were linked primarily with college-level classes, and at Houston, the lowest level of developmental math was linked with the college’s student success class, designed to prepare students for the demands of college. A total of 1,034 students at Queensborough and 1,273 students at Houston entered the study between 2007 and 2009. The key findings presented in this report are:

• Both Queensborough and Houston began by implementing a basic model of a one-semester developmental math learning community; the programs strengthened over the course of the demonstration by including more curricular integration and some connections to student support services.
• Learning community students attempted and passed their developmental math class at higher rates at both colleges.
• In the semesters following students’ participation in the program, impacts on developmental math progress were far less evident. By the end of the study period (three semesters total at Queensborough and two at Houston), control group members at both colleges had largely caught up with learning community students in the developmental math sequence.
• On average, neither college’s learning communities program had an impact on persistence in college or cumulative credits earned.

With these results, a pattern is beginning to emerge in the experimental research on learning communities: Linked classes can have an impact on students’ achievement during the program semester, but this effect diminishes over time. However, a fuller understanding will be gained as findings are released from the remaining three colleges in the demonstration. A final project synthesis report, including further follow-up, will be published in 2012.