Students with one-on-one tutoring outperform peers by two standard deviations, but individualized instruction is economically unfeasible at scale. This session presents an AI-driven “Choose Your Own Adventure” framework. Attendees will explore design blueprints, experience the framework firsthand, and receive adaptable implementation tools for immediate application in asynchronous courses.

The Challenge: Personalization at Scale in Asynchronous Environments

In 1984, educational psychologist Benjamin Bloom published findings that fundamentally challenged traditional education: students receiving one-on-one tutoring with mastery learning outperform conventionally taught peers by two standard deviations. This “2 Sigma Problem” revealed that individualized instruction produces dramatic learning gains, but remains economically impossible to replicate at scale. For forty years, educators have searched for feasible solutions to this challenge.

Today, three converging forces make this problem more urgent than ever. First, students increasingly expect personalization in their educational experiences that mirrors what they encounter in their digital lives, such as customized newsfeeds, algorithm-driven content recommendations, and interfaces that adapt to individual preferences. Second, employers across sectors demand graduates with AI literacy and the ability to effectively leverage artificial intelligence tools. Third, large language models have reached a level of sophistication that allows them to provide individualized feedback, adapt to student responses, and guide learners through complex material. These are capabilities that closely mirror aspects of human tutoring.

For fully asynchronous online courses, this challenge intensifies. Without real-time interaction, students often experience delayed feedback, limited personalization, and a sense of isolation from instructional support. Traditional solutions (e.g., discussion boards, email correspondence, pre-recorded feedback) cannot replicate the immediacy and responsiveness of tutoring.

This session addresses a fundamental question: Can AI-driven learning pathways move us closer to solving Bloom’s challenge in asynchronous environments while simultaneously developing the AI literacy skills students need for their futures?

The Framework: Choose Your Own Adventure Learning Meets Just-in-Time Teaching

This presentation introduces an instructional design model that merges two proven pedagogical approaches: the narrative engagement of Choose Your Own Adventure pathways with the adaptive responsiveness of Just-in-Time Teaching (JiTT). Students progress toward the same learning objectives but select their own pathways based on learning modality preferences, application contexts, and demonstration formats. Rather than following a single prescribed route, learners make 3-5 meaningful decisions throughout a course module that customize their educational journey.

The model integrates AI as an adaptive tutor within existing Learning Management Systems, functioning as an “always-available” electronic performance support system. Using pre-engineered prompts, AI provides just-in-time feedback at decision points, offers scaffolded support, evaluates student work against rubric criteria, and guides learners toward next steps. This approach maintains instructional rigor while offering the individualization that research suggests significantly improves outcomes.

Developed collaboratively by instructional designers, subject matter experts, and content reviewers within a mature online learning environment, this framework represents a systematically designed approach rather than ad-hoc AI experimentation. The design process ensures alignment with quality standards, accessibility requirements, and evidence-based online learning principles.

Key Design Principles:

Asynchronous-First Architecture with JiTT Integration: The framework leverages AI’s 24/7 availability to provide immediate feedback and guidance in fully asynchronous courses where students never meet synchronously. Drawing from Just-in-Time Teaching principles, the system analyzes student responses in real-time and adjusts support accordingly. This provides the “on-demand” requirement that characterizes effective JiTT implementation. Students progress through personalized pathways at their own pace, receiving tutoring-style support regardless of time zone or schedule. This addresses one of asynchronous learning’s core challenges: the delay between student work submission and instructor feedback.

Multiple Pathways, Consistent Outcomes: Students can achieve the same learning objective through different routes. For example, learners might choose to demonstrate understanding through written analysis, oral presentation, or visual representation. These are all assessed against the same competency criteria, but honor different strengths and preferences. This choice architecture mirrors the “interactive engagement” and “active learning” principles that research demonstrates significantly improve student outcomes.

Pre-Engineered Prompt Architecture: Rather than requiring students or instructors to engineer prompts in real-time, the framework provides template-based prompts that specify the AI’s role, establish constraints, define output formats, and ensure consistency across student experiences. This “conversational bot” approach, functioning as an electronic performance support system, reduces the technical barrier to implementation while maintaining quality control and allowing replication across courses and institutions.

Formative Assessment Integration: At each decision point, AI evaluates student work using rubric-based criteria, provides specific feedback for improvement, and determines readiness for advancement. This creates continuous feedback loops that mirror the checking-for-understanding that occurs in one-on-one tutoring. The system replaces the traditional asynchronous pattern of submit-wait-receive feedback with immediate, personalized responses, which is a key advantage of bot-enhanced JiTT over traditional methods in terms of speed, scale, and personalization.

Metacognitive Reflection and AI Literacy Development: Students complete brief reflections on their pathway choices and how AI supported their learning, developing both self-awareness about their learning preferences and critical thinking about AI’s role in education. This explicit AI literacy component ensures students develop skills for evaluating and leveraging AI tools effectively.

Dr. Brent Moore

Chair, Behavioral & Health Sciences (CAPS) at Indiana Wesleyan University-National & Global

Dr. Rick Bartlett

Assistant Director of Faculty Engagement at Indiana Wesleyan University-National & Global