Regular and Substantive Interaction (RSI) is the standard US federal distance-education rules use to separate online courses from correspondence study: instructors must initiate regular, substantive academic interaction with their students. For video-heavy courses, meeting it means moving instructor engagement into the video itself — and keeping time-stamped evidence that it happened.
Under US federal distance-education requirements, a course counts as distance education — rather than correspondence study — when instructors engage in regular and substantive interaction with enrolled students. In practice that breaks into three tests. The interaction must be substantive: direct instruction, assessment or feedback on coursework, answering academic questions, or otherwise engaging students in the substance of the discipline. It must be regular: scheduled and predictable, with instructors monitoring engagement and reaching out proactively rather than waiting for a student to email. And it must be instructor-initiated: the institution cannot rely on students to trigger every exchange.
How accreditors and federal reviewers read those tests continues to evolve, so treat this page as an operational guide and confirm current requirements with your compliance counsel and accreditor before you finalize policy or timelines. What does not change is the practical burden: you need real interaction, on a rhythm, started by instructors — and you need to be able to show it after the fact.
Most online programs lean heavily on recorded video, and recorded video by itself is the textbook example of what RSI is not. A student streaming a lecture alone is interacting with a file, not a faculty member. Watch-time dashboards can prove the video played; they cannot prove anyone taught. If the only human contact in a course is a discussion board nobody visits and an inbox students rarely use, the course starts to look like correspondence study no matter how good the lectures are.
That gap carries real consequences for program classification and federal aid eligibility — stakes your counsel can size for your institution. The fix is not to abandon video. It is to make the video itself the place where interaction happens, so the medium your students already spend their hours in becomes the medium that generates your evidence.
Annoto adds a collaboration and assessment layer on top of the video player your program already uses — Panopto, Kaltura, YouTube, Vimeo, or Annoto's native hosting — and runs inside your LMS. In-video discussion sits on the video timeline: students ask questions at the exact second that confused them, classmates and instructors answer in context, and every exchange is recorded with a timestamp, an author, and a role.
Because each comment, reply, and answer is anchored to a moment in a specific course video, the resulting record is precisely the kind of evidence RSI reviews ask for: who interacted, about what academic content, when, and who started it.
RSI cares about who opens the exchange. With timeline discussion, instructors do not have to wait for questions: they can seed prompts at specific timestamps — a challenge at minute 14 where cohorts historically get lost, a reflection question before a worked example — and reply to student threads in the same context. Each instructor-initiated prompt and each substantive reply is logged where it happened, not buried in a separate forum.
Substantive interaction includes assessing and giving feedback on coursework, and graded activity leaves the cleanest trail. Annoto's in-video quizzes pause the video at chosen moments, check understanding, and sync results to the LMS gradebook over LTI 1.3 — so the assessment record lives in the system your registrar and reviewers already trust.
Video assignments extend the same loop: students submit their own recordings, instructors leave time-stamped feedback on the exact moments that need work, and structured peer review adds documented student-to-student engagement on top. Together these produce graded, instructor-touched artifacts every module — not just view counts.
"Regular" is a claim about cadence, so it needs longitudinal data. Annoto's per-learner engagement analytics show who watched, who participated, who asked and answered questions, and how that changed week over week. Instructors can spot disengaged students early and reach out — which is itself instructor-initiated interaction when it is logged. Program leads get a course-level view of whether interaction is actually happening on the promised schedule, before an auditor asks.
Review rubrics differ by accreditor, but the recurring themes are consistent: a predictable, published schedule of instructor touchpoints; interaction that is academic in substance, not just "great job!" encouragement; clear signs the instructor initiates rather than only reacts; and an evidence trail tied to identifiable course activity and identifiable students.
The common failure patterns are just as consistent: auto-graded quizzes with no instructor involvement, an announcements-only instructor presence, and forums that show activity in week one and silence afterward. Time-stamped in-video interaction addresses all three because the instructor's substantive presence is woven through the actual instruction.
First, map interaction touchpoints per module before the term starts: which videos carry seeded instructor prompts, which carry graded quizzes, and when instructors will hold discussion time. Second, publish the cadence to students — predictability is half of "regular."
Third, commit instructors to substantive, time-stamped replies within a stated window, and grade at least one in-video activity per module through the LTI gradebook sync. Fourth, review engagement analytics weekly, contact learners who have gone quiet, and log the outreach. Finally, export the evidence every term rather than reconstructing it under audit pressure. For a fuller walkthrough of this workflow, see RSI made practical.
When a reviewer asks for proof, Annoto lets you produce it instead of promising it: discussion records with timestamps, authors, and roles; instructor reply activity; quiz and assignment outcomes synced to the gradebook; and per-learner engagement reports across the term. Programs assemble these into a term-by-term evidence packet mapped to their accreditor's expectations — the approach covered step by step in Prove RSI compliance, part of our broader compliance overview.
No. Watching is consumption, not interaction — a completed view involves no instructor and no exchange, which is why watch-time alone cannot carry an RSI case. Watch data still matters as monitoring evidence: it shows instructors tracking engagement and triggering outreach. The interaction itself has to come from instructor activity, such as time-stamped prompts, substantive replies, and graded feedback on the video.
Annoto can export time-stamped in-video discussion records including authors and roles, instructor reply and prompt activity, in-video quiz and video-assignment results synced to the LMS gradebook via LTI 1.3, and per-learner engagement analytics over time. Together they document the substantive, regular, instructor-initiated pattern reviewers ask about — organized by course and by term so your team can hand over a packet, not a scavenger hunt.
Book a demo and see the time-stamped discussion, graded in-video interaction, and exportable per-learner reports live on your own course video.
Book A Demo