Sleep Curriculum Study: Consent, Design, and Data Requests
A suburban middle school (grades 6–8; 1,050 students) wants to test a new eight-week sleep-hygiene curriculum. The district wellness coordinator proposes flipping a coin for each of 30 homerooms to assign the curriculum or the existing health lessons, then comparing anonymous pre/post sleep surveys and quarterly tardy records. The survey asks about bedtimes, caffeine, and screen use; 14% of students are English learners; 12 teachers have volunteered to deliver lessons. A parent objects that her child was never asked to consent. The principal supports the study because tardiness rose 18% last year. The coordinator has no extra staff, wants results by June, and the district receives federal education funds but has no IRB. A reporter has requested homeroom-level tardy data. The coalition must decide whether and how to proceed.
1. Identify the approvals and consent processes the coordinator must complete before collecting data, and explain how parental permission and student assent should be handled for the surveys.
Read the worked answer
Several approvals are needed before any data collection. Because the district receives federal education funds and the activity is designed to produce generalizable knowledge, the coordinator should obtain review by an IRB-often through a partner university or a hired external IRB-unless the district's own review formally determines the activity is exempt. The survey asks minors about sleep, caffeine, and screen use; even if minimal risk, it is not automatically exempt merely because it is school-based.
Parental permission and student assent should both be sought. Active written parental permission is safest for a survey with identifiable homeroom data; passive consent may be acceptable only if the IRB approves it, the opt-out is clearly explained in multiple languages, and participation is truly voluntary. Students should be told they can skip questions or stop. The coordinator must not condition grades or services on participation. Under FERPA, tardy records are education records; the district should use them only under an approved data-sharing plan, de-identify them before analysis, and avoid releasing small cells. The principal should confirm state minor-consent and survey laws before proceeding.
2. Evaluate the proposed cluster-randomized design. What threats to internal validity and feasibility should the coalition anticipate, and what changes would strengthen it?
Read the worked answer
Cluster randomization by homeroom is defensible because the curriculum is delivered to whole classes, but 30 clusters is small and coin-flip allocation can produce imbalanced groups. Homerooms differ by teacher, grade level, and schedule, so the coalition should pair-match homerooms on grade and prior tardiness, then randomize within pairs. Contamination is possible because students and teachers talk across homerooms; measure exposure in the control condition and avoid claiming a pure no-treatment comparison.
Outcomes need a clear definition. Tardy records may reflect bus delays, weather, and attendance policy changes, not sleep. Anonymous pre/post surveys cannot be linked at the student level; use a confidential code or aggregate at homeroom level. Volunteer teachers may deliver lessons with uneven fidelity, so provide a scripted lesson, observation checklist, and attendance logs. The study is underpowered for small effects. Report confidence intervals, pre-specify the primary outcome, and avoid overinterpreting a single semester. A stronger design would add a delayed-start comparison and collect fidelity and student-reported sleep duration.
3. The reporter requests homeroom-level tardy data. How should the district respond under FERPA and research ethics?
Read the worked answer
Tardy data are education records protected by FERPA. The reporter is not entitled to homeroom-level counts if those counts could identify students, especially where a homeroom has few students or a distinctive attendance pattern. The district should route the request to its records officer, release only aggregate results after the study, and apply a small-cell suppression rule-for example, do not report cells with fewer than 10 students. It should not release student names, ID numbers, or any combination that allows re-identification. If the district wants to share findings, it can report school-wide or grade-wide percentages, describe the study methods, and explain limitations. Research consent forms should state whether aggregate results may be shared. The district should also avoid letting media coverage pressure teachers or families into participation. If the reporter seeks interviews, use the district communications process and protect student privacy.