Faculty Guide

Overview

The SAIL Challenge is a structured learning experience that develops students' ability to work with AI while maintaining critical judgment. It can be embedded in any course where students analyze cases, solve problems, or make recommendations.

The Challenge takes approximately 90-120 minutes of student work and can be:

• A standalone assignment
• Integrated into an existing case analysis assignment
• Part of a larger project
• A credentialed micro-learning experience

Pedagogical Foundation

The SAIL Challenge design is grounded in over 25 years of learning research, including two major meta-analyses:

• Comparison Creates Learning (Schwartz & Bransford, 1998; Alfieri et al., 2013): When learners compare contrasting cases, they notice critical distinctions and develop "differentiated knowledge structures" that prepare them to learn deeply. A meta-analysis of 57 experiments found case comparison produces significantly greater learning (d = 0.50), with effects largest when explanatory principles emerge after comparison (d = 1.18) — the exact sequence the SAIL Challenge uses.
• Productive Failure (Sinha & Kapur, 2021): A meta-analysis of 53 studies found that problem-solving before instruction significantly outperforms instruction-first approaches (g = 0.36). When implemented with high fidelity, effects reach g = 0.37–0.58. Phase 1 creates productive struggle that prepares students to learn from the AI comparison.
• Desirable Difficulties (Bjork & Bjork, 2011): Phase 1 creates productive struggle by requiring students to think before AI assists. Conditions that slow immediate performance often enhance long-term retention and transfer.
• Metacognition (Flavell, 1979): Phase 3's reflection requirements develop students' awareness and regulation of their own thinking processes.
• Transfer of Learning (Perkins & Salomon, 1988): The constant structure with varying content enables both automatic (low-road) and mindful (high-road) transfer across contexts.
• Critical Thinking (Facione, 1990): The Challenge develops both cognitive skills and dispositional qualities that constitute "purposeful, self-regulatory judgment."

Adapting for Your Discipline

The SAIL Challenge works with any case, problem, or scenario that requires analysis and recommendation. The key is choosing material that:

• Has no obvious right answer — Reasonable people can disagree
• Requires judgment, not just calculation — AI can help but cannot decide
• Has both quantitative and qualitative dimensions — Creates interesting comparison between human and AI analysis
• Connects to course learning objectives — Reinforces what you're teaching

Implementation Options

Option A: Single Class Session

Complete all three phases in one extended class (2-3 hours) or across two regular sessions.

• Phase 1: 30-40 minutes in class (no devices or AI)
• Phase 2: 30-45 minutes with AI access
• Phase 3: 20-30 minutes writing judgment memo
• Debrief discussion: 20-30 minutes

Option B: Homework Assignment

Assign as take-home work with phase submissions.

• Phase 1 due first (e.g., Tuesday midnight)
• Phase 2 and 3 due later (e.g., Thursday midnight)
• In-class debrief discussion

Option C: Integrated with Existing Case

Apply the three-phase structure to a case you already use.

• Modify your existing assignment instructions to include the three phases
• Add the Phase 2 AI Evaluation Log requirement
• Revise rubric to include AI evaluation dimensions

Ensuring Academic Integrity

The Challenge's validity depends on authentic Phase 1 work. Here are strategies to protect integrity:

Structural Safeguards

• Sequential submission: Require Phase 1 submission before releasing Phase 2 instructions
• In-class Phase 1: Have students complete Phase 1 in class without devices
• Timestamped submissions: Use LMS features to timestamp when Phase 1 was submitted
• Comparison analysis: If Phase 1 and Phase 2 are nearly identical, that's a red flag

Detection Indicators

• Phase 1 writing style dramatically different from student's other work
• Phase 2 evaluation shows no meaningful differences between student analysis and AI analysis
• Phase 3 cannot articulate where they diverged from AI (because there was no divergence)
• Phase 1 contains hallmarks of AI writing (certain phrases, structures, or breadth of reference)

Facilitating the Debrief

The post-Challenge discussion is where much learning is consolidated. Consider these discussion prompts:

Common Student Realizations

Students often discover:

• AI is better at breadth (considering many factors) but weaker at depth (nuanced judgment)
• AI often sounds confident even when wrong
• Their own thinking had gaps they didn't notice until seeing AI's perspective
• The hardest part is deciding when to trust AI, not how to use it
• Writing the judgment memo forced them to articulate reasoning they hadn't made explicit

Assessment Considerations

The provided rubric can be used as-is or adapted. Key assessment principles:

• Assess process, not just product: The quality of thinking matters more than the "correctness" of the recommendation
• Value honest evaluation: A student who identifies AI's limitations shows more learning than one who accepts AI uncritically
• Reward metacognition: Students who can articulate what they learned about their own thinking demonstrate deeper learning
• Weight Phase 3 heavily: The Judgment Memo is where students demonstrate integrated learning

Credential Integration

The SAIL Challenge can be part of a credentialing pathway. Students who complete the Challenge at a "Proficient" level or above earn a digital credential demonstrating:

• Ability to analyze problems independently
• Skill in critically evaluating AI output
• Capacity to integrate human judgment with AI assistance
• Professional communication of reasoning and decisions

Contact the SAIL Collaborative for information on credential integration.

Quick Start Checklist

Support and Resources

The SAIL Collaborative is available to support faculty implementing the Challenge:

• Case development assistance
• Rubric customization
• LMS setup guidance
• Credential integration
• Research collaboration on student outcomes

Contact: Hasan Arslan, PhD — harslan@suffolk.edu