Preventing Student Dropouts in the Pandemic

© Is­abel Cor­thi­er

Stu­dent dropouts con­sti­tute a sig­nif­i­cant con­cern in de­vel­op­ing coun­tries; even be­fore the pan­dem­ic, one out of three Brazil­ian stu­dents dropped out be­fore grad­u­at­ing high school. School clo­sures in the con­text of COVID-19 have been shown to mag­ni­fy that prob­lem, with at least sev­en mil­lion ad­di­tion­al dropouts world­wide in 2021. De­spite ef­forts from gov­ern­ments world­wide to mit­i­gate learn­ing gaps by the time in-per­son class­es re­turn, in­ter­ven­tions to mo­ti­vate stu­dents to re­main in school un­til then have been over­looked.

The Sus­tain­able De­vel­op­ment Goals of this Pro­ject
quality education

Re­cent ev­i­dence shows that, in de­vel­op­ing coun­tries, be­hav­ioral nudges – typ­i­cal­ly sent through text mes­sages to par­ents’ cell phones – have the po­ten­tial to not only sig­nif­i­cant­ly im­prove learn­ing out­comes, but also to dras­ti­cal­ly de­crease dropouts be­fore the pan­dem­ic. Hav­ing said that, it would be sur­pris­ing if those ef­fects repli­cat­ed dur­ing school clo­sures be­cause such nudges typ­i­cal­ly work by in­duc­ing par­ents to show up in school to a greater ex­tent and mon­i­tor teach­ers more close­ly – a mech­a­nism that might not be mean­ing­ful in the ab­sence of in-per­son class­es.

This project stud­ies whether be­hav­ioral nudges to mo­ti­vate high-school stu­dents to stay en­gaged with school ac­tiv­i­ties dur­ing the pan­dem­ic sub­stan­tial­ly de­creased dropout risk dur­ing school clo­sures in Brazil. In part­ner­ship with In­sti­tu­to Son­ho Grande and the Goiás State Sec­re­tari­at of Ed­u­ca­tion in the con­text of their full-time high school pro­gram (En­si­no Mé­dio em Tem­po In­te­gral), 12,056 high-school stu­dents across 57 pub­lic schools re­ceived nudges, while oth­er 6,200 high-school stu­dents across 30 pub­lic schools re­ceived no nudges or oth­er text mes­sages from their schools over that pe­ri­od.

We found that nudges de­creased dropout risk by around 26% over the course of the school year. Ef­fects in­creased with ex­po­sure and were con­cen­trat­ed among stu­dents at the high­est risk of dropouts. Nudges only worked when sent di­rect­ly to stu­dents’ phones and in schools that al­ready of­fered on­line aca­d­e­m­ic ac­tiv­i­ties pri­or to the pan­dem­ic. Fram­ing con­tent in terms of the up­side of grad­u­at­ing high school led to high­er im­pacts than fram­ing it in terms of the down­side of drop­ping out. Al­lud­ing peer mo­ti­va­tion to re­turn to in-per­son class­es to lever­age so­cial pres­sure had no ad­di­tion­al ef­fects on dropout risk.

Re­sults show that be­hav­ioral nudges can par­tial­ly mit­i­gate the dra­mat­ic in­crease in stu­dent dropouts dur­ing school clo­sures by keep­ing ado­les­cents mo­ti­vat­ed to stay in school. The pat­terns of het­ero­ge­neous treat­ment ef­fects are con­sis­tent with com­ple­men­tar­i­ties be­tween mo­ti­va­tion and aca­d­e­m­ic in­struc­tion. All in all, our re­sults show­case that in­sights from the sci­ence of ado­les­cent psy­chol­o­gy can be lever­aged to shift de­vel­op­men­tal tra­jec­to­ries at a crit­i­cal junc­ture, but also raise cau­tion against in­dis­crim­i­nate­ly ap­ply­ing be­hav­ioral in­sights de­rived from eval­u­a­tions of sim­i­lar in­ter­ven­tions in con­texts of in-per­son class­es or sta­t­ic de­ci­sion-mak­ing.

The loop: ado­les­cent un­der-ed­u­ca­tion, as mul­ti­ple forces com­bine to push stu­dents out of school dur­ing the pan­dem­ic

Break­ing the loop: mo­ti­vat­ing ado­les­cents to stay en­gaged with re­mote learn­ing ac­tiv­i­ties through be­hav­ioral nudges sent di­rect­ly to their cell phone

Pre­vent­ing Stu­dent Dropouts in the Pan­dem­ic

Stu­dents at­ten­dance on dis­tance learn­ing has de­creased, while dropout rates have in­creased. In this study in Brazil, we aim at in­creas­ing the school en­gage­ment of stu­dents and par­ents by send­ing them text nudges.

  • Sta­tusCon­clud­ed
  • Coun­tryBrazil
  • Part­nersIn­sti­tu­to Son­ho Grande, Goiás State Sec­re­tari­at of Ed­u­ca­tion, MOV­VA
  • Time­line2020
  • Study TypeRan­dom­ized con­trolled tri­al
  • Sam­ple Size12,056 stu­dents

Re­search Team

Julien Christen