Research Literacy

Purpose

AI agents tend to execute analysis steps immediately without planning or justification. In research, every analysis decision needs a rationale grounded in theory, design, and data characteristics. This skill encodes the basic scientific thinking that should precede any domain-specific action.

A competent programmer without research training will typically: (a) pick a familiar method rather than the appropriate one, (b) skip assumption checks, (c) interpret results without considering alternative explanations, and (d) make undisclosed analytic choices that inflate false positive rates. This skill exists to prevent all four failure modes.

When to Use

• Before or alongside any domain-specific skill from this project (e.g., before running an ERP analysis, first formulate the research question and justify the method).
• Standalone when planning a study, reviewing an analysis pipeline, or interpreting results.
• Whenever an analysis involves researcher degrees of freedom — choices that could have been made differently and would affect the outcome.

⚠️ Verification Notice