🎓 SPL Pedagogies Gallery

15 Research-Based Intelligent Tutoring System Pedagogies

Pedagogies

+0.30-0.79

Effect Sizes (SD)

100%

LLM-Powered

∞

Learning Paths

Transforming SPL Tutoring

Welcome to the SPL Pedagogies Gallery – a showcase of 15 evidence-based intelligent tutoring system pedagogies that transform the Socratic Playground from a conversational tutor into a comprehensive adaptive learning companion. Each pedagogy is grounded in decades of learning sciences research with documented effect sizes ranging from +0.30 to +0.79 standard deviations. Explore research-backed tutoring strategies that adapt to individual learners, from Socratic questioning to productive failure, peer tutoring to metacognitive scaffolding.

📖 Comprehensive Pedagogy Guide

Deep dive into learning theories, domain specificity, research citations, and best practices for each pedagogy

Explore Pedagogy Guide & Learning Theories

🎯 Pedagogies by Category

Choose pedagogies based on your learning goals and content type

🧠 Conceptual Depth

4 Pedagogies

⚙️ Procedural Skills

3 Pedagogies

#5 Cognitive Apprenticeship (+0.76 SD) #7 Example-Based Learning (+0.57 SD) #9 Model-Tracing Tutoring (+0.76 SD)

🔥 Motivation & Engagement

2 Pedagogies

#4 Teachable Agents (+0.50 SD) #15 Interleaved Practice (+0.42 SD)

🤝 Collaborative Learning

2 Pedagogies

#3 Peer Tutoring (+0.40 SD) #6 Reciprocal Teaching (+0.60 SD)

🧩 Metacognition

3 Pedagogies

#11 Self-Explanation (+0.61 SD) #12 Metacognitive Scaffolding (+0.53 SD) #14 Elaborative Interrogation (+0.50 SD)

🎨 Open-Ended Design

1 Pedagogy

#8 Constraint-Based Modeling (+0.71 SD)

⚡ Strategic Adaptation

1 Pedagogy

#13 Adaptive Feedback Timing (+0.30 SD)

🧠 All 15 SPL Pedagogies

Pedagogy 01

Socratic Method

Guided discovery through strategic questioning. Tutor never directly provides answers but leads learner to construct knowledge through inquiry. 70-80% of tutor turns are questions.

✅ Interactive Demo Medium +0.79 SD

Pedagogy 02

Productive Failure

Learners attempt complex problems BEFORE receiving instruction. Struggle (10-20 min) → Consolidation. 60-80% failure rate expected and productive for deep learning.

✅ Interactive Demo Hard +0.20 SD Transfer

Pedagogy 03

Peer Tutoring

Students take turns being tutor and tutee, learning by teaching and receiving explanations. AI monitors for errors and prompts deeper explanations. Reciprocal roles every 5-10 minutes.

✅ Interactive Demo Medium +0.40 SD

Pedagogy 04

Teachable Agents

Student teaches virtual AI agent that makes plausible mistakes based on teaching quality. Agent's quiz performance reflects teaching effectiveness. Protégé effect increases motivation.

✅ Interactive Demo Easy +0.50 SD

Pedagogy 05

Cognitive Apprenticeship

Expert models problem-solving (thinking aloud) → Scaffolding → Fading → Independence. Gradual release: I do → We do → You do (with support) → You do (alone).

🔄 Specification Ready Hard +0.76 SD

Pedagogy 06

Reciprocal Teaching

Students lead discussion using 4 strategies: Clarify → Question → Summarize → Predict. Role rotation with AI as strategy coach. Builds metacognitive reading comprehension.

🔄 Specification Ready Medium +0.60 SD

Pedagogy 07

Example-Based Learning

Study worked examples → Self-explanation prompts → Faded examples → Practice. Reduces cognitive load for novices. Emphasis on explaining WHY each step works.

🔄 Specification Ready Easy +0.57 SD

Pedagogy 08

Constraint-Based Modeling

Open-ended problems with constraint violation detection. AI monitors for what NOT to do rather than prescribing correct path. Multiple valid solutions encouraged.

🔄 Specification Ready Very Hard +0.71 SD

Pedagogy 09

Model-Tracing Tutoring

AI maintains fine-grained cognitive model of solution paths. Traces student steps and intervenes immediately after errors with minimal hints. Multiple correct paths supported.

🔄 Specification Ready Medium +0.76 SD

Pedagogy 10

Analogical Reasoning

Learn new concept by comparing to familiar analog. Source analog → Mapping → Transfer → Limits. Focus on structural similarity. Discuss where analogy breaks down.

🔄 Specification Ready Medium +0.50 SD

Pedagogy 11

Self-Explanation

Students explain concepts/solutions to themselves, generating inferences that fill knowledge gaps. AI prompts deeper explanation if superficial (4 levels: paraphrase → elaboration → inference → principled).

🔄 Specification Ready Medium +0.61 SD

Pedagogy 12

Metacognitive Scaffolding

AI prompts planning, monitoring, and evaluation. Pre-task (goals/strategy) → During-task (progress check) → Post-task (self-assessment). Develops self-regulated learning skills.

🔄 Specification Ready Hard +0.53 SD

Pedagogy 13

Adaptive Feedback Timing

Strategically times feedback: Immediate for procedural errors, delayed for conceptual understanding. Decision based on error type and learning goal. Optimizes desirable difficulties.

🔄 Specification Ready Medium +0.30 SD

Pedagogy 14

Elaborative Interrogation

Students answer "why" questions about facts and concepts. 70-80% of prompts start with "Why." Promotes deep processing, causal reasoning, and connection to prior knowledge.

🔄 Specification Ready Easy +0.50 SD

Pedagogy 15

Interleaved Practice

Mix problem types during practice (not blocked). No more than 2 consecutive same-type problems. Initially harder but superior long-term retention and discrimination. +0.42 SD on delayed tests.

🔄 Specification Ready Easy +0.42 SD

🗺️ Implementation Roadmap

Phase 1: Core Pedagogies (Months 1-3)

Already Implemented: Socratic Method, Adaptive Feedback (basic), Self-Explanation
Quick Wins: Elaborative Interrogation, Teachable Agents, Interleaved Practice

Phase 2: Advanced Pedagogies (Months 4-8)

Cognitive Apprenticeship (modeling mode), Example-Based Learning (worked examples library), Analogical Reasoning (analogy database), Metacognitive Scaffolding (planning/monitoring prompts)

Phase 3: Research-Grade (Months 9-12)

Productive Failure (delayed instruction mode), Peer Tutoring (multi-user pairing), Reciprocal Teaching (4 Cs framework), Constraint-Based Modeling, Model-Tracing (cognitive model)

🎯 Why These Pedagogies Matter

Research-Based: All 15 pedagogies have documented effect sizes from +0.30 to +0.79 standard deviations.
Diverse Approaches: Conceptual depth, procedural skills, motivation, collaboration, metacognition.
Adaptive: AI selects optimal pedagogy based on learner profile, content type, and learning goals.
Engagement: 15 varied tutoring strategies prevent monotony and match individual learning preferences.
Transfer: Strategies like Productive Failure (+0.20 SD) specifically target transfer to novel situations.

📖 Complete Documentation: See SPL_Variety_Form.md for full specifications, research citations, implementation patterns, and integration with the 100-agent ITS ecosystem.

Created with 🧠 by UALS Team | Intelligent Tutoring Systems for Adaptive Learning