Substantive Interaction by Design

What RSI Actually Requires

In 2021, the Department of Education clarified what distinguishes distance education from correspondence courses: Regular and Substantive Interaction between instructors and students. The stakes are significant—courses lacking RSI may be reclassified as correspondence education, affecting financial aid eligibility and accreditation standing.

But the regulations aren't arbitrary bureaucracy. They codify what decades of research already established: meaningful instructor presence drives student success. When instructors actively engage—facilitating discussion, providing substantive feedback, responding to questions about content—students learn more, persist longer, and report higher satisfaction.

The question for institutions isn't simply "Can we document compliance?" It's more fundamental: Are our tools designed to produce substantive interaction, or just to capture evidence that something happened?

There's a difference between a platform that makes compliance documentation convenient and one that makes substantive interaction natural. VoiceThread was designed for the latter.

The Five Forms of Substantive Interaction

The 2021 regulations recognize five forms of substantive interaction. Courses must demonstrate at least two. VoiceThread directly enables four of the five—and was purpose-built for the one regulators call out explicitly: facilitating group discussion.

This is VoiceThread's primary purpose. Unlike text-based discussion boards where instructors post prompts and return to grade, VoiceThread creates a seminar-style environment where instructor facilitation is visible, expected, and natural.

• Instructors frame discussions with voice or video, establishing presence from the start
• A "Student Gallery" feature set enables students to give structured and tracked feedback on peer work
• Mid-discussion contributions connect student ideas, redirect when needed, and deepen inquiry
• The visual organization shows where instructor engagement has occurred—and where it hasn't
• Voice and video make substantive facilitation easier than typing, encouraging genuine dialogue rather than perfunctory replies

The regulations explicitly distinguish facilitation from simply "assigning discussion posts and grading participation." VoiceThread's design makes that distinction operational.

"[The instructor] felt very interactive... she left comments and it felt like she was really there with us."

— Online student, University of Kansas

Effective feedback does more than evaluate—it teaches. The regulations require feedback that "communicates accomplishments, areas for improvement, and concrete suggestions for progress." Brief comments like "Good job!" don't qualify.

Voice and video feedback naturally tends toward substance. It's awkward to record a 30-second video that says only "nice work"—the format demands elaboration. Instructors explain their thinking, point to specific moments in student work, and convey tone and encouragement that text cannot capture.

Research confirms students perceive voice feedback as more caring and more detailed, even when conveying similar content to written comments. The medium shapes the message.

"VoiceThread helps my online students feel my constant presence in our course as well as my commitment to their progress as individual learners with unique needs."

— Dr. Abigail Alexander, Assistant Professor of French

When an instructor answers one student's question well, that explanation can benefit everyone. VoiceThread transforms private email exchanges into visible learning resources.

Questions and answers remain connected to relevant content rather than buried in chronological threads. Voice and video responses can demonstrate and elaborate in ways text cannot. Complex concepts can be explained with annotation, pointing, and visual demonstration—not just words.

While VoiceThread is primarily asynchronous, it extends direct instruction beyond the synchronous session. Instructor voice and video presence can surround content throughout the learning arc—contextualizing, highlighting, and connecting—rather than appearing only at the beginning.

This transforms content delivery into ongoing instruction, with instructor presence woven through rather than preceding student engagement.

Why Design Philosophy Matters

Many platforms now claim RSI support. The difference isn't feature lists—it's underlying architecture.

Voice-first, not text-primary. Speaking is humanity's oldest and most natural mode of complex communication. When voice is the expected default rather than an optional add-on, contributions become longer, more nuanced, and more substantive. Text-based platforms produce text-length thinking.

Simultaneity, not sequence. VoiceThread enables speaking while pointing, annotating while explaining. This isn't a minor interface detail—it engages fundamentally different cognitive processes.

Research from the University of Chicago demonstrates that when learners express information through speech and gesture simultaneously, they reveal understanding that neither mode captures alone. A landmark study found that simultaneous speech-plus-gesture significantly outperformed sequential presentation; sequential approaches showed no advantage over speech alone. The researchers concluded that "simultaneity is an important ingredient in gesture's ability to facilitate learning."

When students speak while drawing on content, and when instructors provide feedback while pointing to specific moments in student work, both engage these integration mechanisms. Sequential approaches—type a comment, then record audio, then add an annotation—don't produce the same cognitive benefits.

Presence, not just content. The goal isn't documenting that interaction occurred. It's creating the conditions where instructors and students feel genuinely present to each other despite separation in space and time. Research consistently shows that voice and video substantially enhance social presence compared to text-only discussion, with large effect sizes for feelings of connection, community, and "being part of a class."

"The feedback that I received from students is that they liked getting to know their colleagues better and felt more connected to the classroom, even though it was a distance class."

— Dr. Imelda Reyes, Clinical Associate Professor, Emory Nell Hodgson Woodruff School of Nursing

Assessment Integrity in the AI Era

RSI requirements and academic integrity concerns point in the same direction: toward human presence that can't be automated or fabricated.

RSI requires qualified human instructors substantively engaging with students—not automated feedback systems or AI-generated responses. The regulations explicitly exclude "computer-generated feedback on objective assessments" from counting as substantive interaction.

Simultaneously, generative AI has made text production trivially easy. Written assignments that once demonstrated understanding can now be generated without it. Institutions face a fundamental question about what forms of student expression authentically reveal learning.

Multimodal expression—speaking while annotating, demonstrating while explaining—resists fabrication in ways text cannot. Students can't paste AI-generated content into a voice recording. The simultaneity of speaking while pointing creates temporal pressure that reveals authentic cognitive processes. The embodied nature of multimodal demonstration manifests understanding rather than merely describing it.

VoiceThread wasn't designed as an "AI-proof" assessment tool. But its multimodal architecture produces exactly the kind of performative, embodied demonstration that emerging research identifies as resistant to fabrication—while its voice-first design keeps human instructors at the center of the learning experience.

The Research Foundation

VoiceThread's approach rests on converging evidence from cognitive science, online learning research, and accessibility studies.