Explore·Research Planning·Augmentation·Developing·EXP-004

Research Plan Critique

Value hypothesis

Reduces the risk of flawed research design by subjecting plans to structured adversarial critique, identifying blind spots and biases before they impact data collection or synthesis.

Quality · Insight

The researcher asks an LLM to review their draft research plan from an adversarial stance, critiquing assumptions, questioning methodology choices, looking for possible issues and potential biases in the study design. Unlike co-creation, the AI here is explicitly a challenger, and the researcher should integrate or rebut the proposed critiques. Recommended practice extends this into a formal bias audit after synthesis, prompting AI to identify which participant segments or perspectives may have been underweighted in the analysis.

Risks in application

Bias Bleed

The bias-checking tool itself carries embedded biases, making the critique potentially unreliable or limited for certain domains.

Shallow Solutions

Critique appears thorough but misses domain-specific, cultural, or contextually specific issues that human expertise would have found.

Expertise that differentiates

Research and Insight

Evaluating which critiques are valid vs. spurious. Knowing when to keep the original, and when challenges merit rework.

Ethical Assessment

Bias detection, fairness evaluation, and recognizing whose perspectives are structurally absent from the research design.

AI Fluency that assures

Task Delegation

Framing the delegation as adversarial requires specifying which dimensions the AI should attack. A vague 'critique this' produces generic objections; a structured brief produces targeted ones.