Tier 1 – Most Versatile & Effective:
CO-STAR (Context, Objective, Style, Tone, Audience, Response Format)
- 🥇 Winner of Singapore’s GPT-4 Competition
- Best for: Content creation, marketing, audience-specific writing
TIDD-EC (Task Type, Instructions, Do, Don’t, Examples, User Content)
- Best for: Precision tasks, data analysis, technical work
RTF (Role, Task, Format)
- Best for: Quick, versatile prompting – the “Swiss Army knife”
Tier 2 – Specialized Excellence:
RISEN (Role, Instructions, Steps, End goal, Narrowing)
- Best for: Complex project breakdown, multi-step processes
RODES (Role, Objective, Details, Examples, Sense Check)
- Best for: Example-driven prompts, template-based work
Chain of Thought
- Best for: Problem-solving, logical reasoning, math
- Simple addition: “Let’s think step-by-step”
Tier 3 – Quick & Focused:
ICIO (Instruction, Context, Input, Output)
- Best for: Data processing, research synthesis
APE (Action, Purpose, Expectation)
- Best for: Simple, focused tasks
CLEAR (Concise, Logical, Explicit, Adaptive, Reflective)
- Best for: Evaluating and improving existing prompts
STAR (Situation, Task, Action, Result)
- Best for: Business problem-solving, case studies
Quick Selection Guide:
For Beginners: Start with RTF or APE
For Content Creation: CO-STAR or RODES
For Data Analysis: TIDD-EC or ICIO
For Problem Solving: Chain of Thought or STAR
For Complex Projects: RISEN or CO-STAR
Key Insights from Research:
- CO-STAR is the most comprehensive and proven framework (competition winner)
- RTF is the most versatile for general use
- Chain of Thought dramatically improves reasoning with minimal effort
- TIDD-EC excels when precision is critical
- Simple frameworks often work better than complex ones for basic tasks