Personal or Personalized? Investigating the Impact of Counselor Outreach Versus Automated and Personalized Messaging on Mitigating Summer Melt Among College-Intending, Low-Income High School Graduates

• In Event: 73.047 - The Not-So-Lazy Days of Summer: Interventions to Improve Achievement and Attainment

Abstract

Background:
Policymakers have largely overlooked the summer after high school as an important time period in students’ transition to college. However, successful college matriculation is contingent on students completing a number of financial and procedural tasks during the summer. Several studies document summer attrition rates, ranging from 10 – 40 percent, among students who had been accepted to and intended to enroll in college as of high school graduation (Authors, 2013; Daugherty, 2011; Schooley, Matthews, & Vossler, 2011).

Recent experimental interventions indicate that students’ postsecondary plans are responsive to additional outreach during the summer. In one set of experiments, college counselors offered students help with financial aid and required paperwork. Students offered additional support were 5 – 14 percentage points more likely to matriculate (Authors, 2012; Authors, 2014). In subsequent experiments, peer mentor outreach or automated and personalized text message outreach have yielded similarly positive effects (Authors, 2013).

Purpose:
It is an open question whether one approach is more effective than the other. Students and their families may appreciate and be responsive to personal outreach from a counselor over the summer. Yet, such outreach is time-consuming for counselors. In contrast, automated and personalized text messaging is a low-cost approach for reaching out to students in order to provide them with information and access to assistance.

We designed and implemented a randomized trial to investigate the comparative efficacy of counselor and automated text outreach. In the first treatment, counselors reached out to randomly selected students to offer them support with their college transition. In the second treatment, students received, ten automated text message reminders of key tasks to complete over the summer. A third group served as the control.

Empirical strategy
We collaborated with a network of five school districts in Texas; uAspire, a college access organization in Boston; and CollegeBound Foundation, a college access organization in Baltimore, to implement the experiment. Our total sample included 9,596 students; 4,063 students were assigned to receive texts, 1,738 were assigned to receive counselor outreach, and 3,795 were assigned to the control.

We capitalize on data from two primary sources. Each site provided student-level demographic and prior academic achievement information and interaction-level records from the text message platform and counselor interaction logs. We obtained student-level college enrollment records from the National Student Clearinghouse (NSC). Given our experimental design, we employ straightforward regression techniques to investigate the impact of both treatments on students’ college outcomes.

Findings:
Both the text and counselor outreach strategies had positive impacts on enrollment, with effects most pronounced for lower-income students and students with less access to quality college information or advising. For instance, in one of our partner sites, low-income students who received the text messages were four percentage points more likely to enroll in college, while students who received counselor outreach were two percentage points more likely to go to college. We find similar magnitude impacts of text- and counselor-outreach, suggesting that text messaging may be a particularly cost effective strategy to increase college enrollment among underrepresented students.

Authors

• Benjamin Castleman, University of Virginia

• Lindsay Coleman Page, University of Pittsburgh