Causal Research Framing

Explore where causality can help with your research. Nirvana helps you clarify your question and prepare a focused brief for further analysis with experts or AI tools.

What decision or scientific question do you care about?

Markdown supported

0 / 2000

Load an example (replaces this worksheet and chat)

What kind of problem is this?

New to this? Learn it in the Portal

Connect your question to the framing

Your current reasoning or uncertainty

Clarify the causal contrast

Each Brief field supports up to 2,000 characters. Longer drafts are kept, but must be shortened before asking Nirvana for edits.

Intended decision or purpose Exposure or intervention Comparator Outcome Population Time horizon Assumptions & uncertainty

Prediction or generation target and intended use Candidate predictors and when they are measured Training environment Expected deployment environment Suspected shifted variable or mechanism How the predictors may relate to the target Deployment data and adaptation constraints Assumptions & uncertainty

Intended decision or scientific purpose Exposure or intervention Proposed mediator Outcome Population Target effect Time ordering Assumptions & uncertainty

Purpose Population or cohort Measure Denominator Breakdowns or strata Time window Data source & limitations

Scientific purpose Candidate variables and measurement timing Observations, interventions and sampling Known ordering and structural constraints Latent variables, measurement and uncertainty

Decision and target population Available actions and current policy Outcome, utility and time horizon Information available when choosing Budget, capacity and other constraints Data and evidence about action effects Assumptions and uncertainty

Purpose and factual case Observed state and outcome Hypothetical intervention and contrast Counterfactual outcome of interest Structural knowledge and case-preserving assumptions Evidence, identifiability and uncertainty

Intended use and population Target and prediction horizon Inputs available at prediction time Training and evaluation data Success criteria and validation plan Leakage, missingness and uncertainty

Where causality can be applied

Choose one bounded next step

Causal Research Brief

DRAFT