Blog: Data, Research & Evaluation

Report Roundup: Does Nudging Work or Doesn’t It? Are Fewer Loans Worse for Students?

Monday, August 5, 2019  
Posted by: Bill DeBaun, Director of Data and Evaluation
Share |

It can be hard to keep up with all of the latest reports. Acknowledging that, here are summaries of three recent interesting pieces of research.

Bad News for Nudging: In “The Remarkable Unresponsiveness of College Students to Nudging And What We Can Learn from It,” Philip Oreopoulos and Uros Petronijevic find in a five-year study of 25,000 students across three college campuses that randomly assigned nudging interventions aimed at improving academic outcomes had no effect. The authors find some benefits on students’ mental health and study habits (coaching interventions increased study time by about two hours per week).

The report partially concludes: “We also find that our coaching interventions had no discernable impact on grades or persistence. Students certainly appreciated receiving follow-up text messages and virtual coaching support after completing an online exercise with advice on how to have a successful year. Coaches made students feel more supported and even happier, but they did not significantly improve performance.” The devil is in the details with nudging in terms of frequency and intensity of interventions, etc., but this is a multi-site study with a large sample over a long timeframe and its failure to find meaningful impacts from nudging is concerning.

But Also Good News for Nudging: It’s not all bad news for nudging this week. A much smaller study relative to the one above (but still large in social science terms) found that in a sample of 1,982 students randomly assigned to an intervention where they gave motivational advice to younger students, those students receiving the intervention saw their grades improve when measured against a control group. This study seems ripe for replication (would we see the same or similar results if the experiment was conducted again), but it gives programs incorporating peer mentorship something to hang their hat on.

Loans and Outcomes at Community Colleges: A new study of students at the Community College of Baltimore County in Maryland finds that not taking out enough in loans can be detrimental, just like taking out too much. Students randomly received a monthlong intervention consisting of eight text messages reminding them that they can choose how much to take out in loans, including opting not to take out the maximum, discussing loan payment amount fluctuations based on disbursement, and informing them about lifetime loan limits, among other topics (see Appendix B at the link). The authors found that students who received the text messages reduced their unsubsidized loan borrowing by 7% (about $200) on average. Students who received the text message also were less likely to earn any credits that semester and more likely to fail a course. They were also more likely (2.5 percentage points) to default on their loans within three years, relative to a control group. This, along with other research on a similar topic offers evidence that not taking out enough loans may contribute to debt with no degree.

Bonus! Persistence and Retention Rate Benchmarks: “The National Student Clearinghouse Research Center is a valuable resource” is a message that Success Digest has hammered home time and again. Let’s do it one more time. They’re out with a new report “Persistence & Retention – 2019” that provides national benchmarks on these outcomes for students who first began college in fall 2017.

Overall, 74% of students persisted anywhere in the fall of 2018, and 62% were retained by the institutions at which they started. Although the persistence rate stayed mostly the same year-over-year, it has dipped by 2.2 percentage points since fall 2009. These metrics are valuable benchmarks for NCAN member programs, and the report disaggregates persistence and retention by enrollment intensity, non-degree-seeking students, student age at college entry, race/ethnicity, and starting institution type.