Leveraging Parents through Low-Cost Technology

middle and high school students across participating schools; 32,000+ text messages sent

US middle and high school students and their parents

Randomised trial automating the extraction of student data from school information systems and pushing weekly text alerts to parents about missed assignments, grades and class absences. Critically - and unlike the mailing trials - this study measured BOTH the behavioural outcome (attendance) and the achievement outcome (state standardized tests), so the two steps can be evaluated separately in the same experiment. Total variable cost $63.

The intervention raised class attendance by 12% and cut course failures by 27%, and increased student retention - but there was NO impact on state test scores. This is the single most informative result in the topic: a randomised intervention that clearly moved attendance and clearly moved grades, and did not move independently measured achievement. Effects were larger for below-median-GPA students and high school students.

Low - randomised. The value of this trial for the hereditarian lens is the dissociation it produces. Course grades (which reward attendance, homework submission and compliance directly) moved 27%; standardized tests (which measure what the student knows) did not move at all. That divergence is exactly what you would expect if the attendance-achievement correlation in observational data is largely about student and family type rather than instructional dose.