Checklist

Exploratory Data Analysis Validation Checklist & Hypothesis Log

Interactive checklist and hypothesis log to capture EDA findings, confirm reproducibility and data quality, and convert observations into prioritized, testable hypotheses with owners and next steps.

Interactive Tool

EDA Validation Checklist & Hypothesis Log

Use this form to record a reproducibility-focused EDA session: verify data quality and provenance, document checks and assumptions, and capture each promising observation as a discrete hypothesis with a test plan, priority, owner and status. Save one hypothesis per submission and repeat as needed. Responses are stored so teams can review, export, and hand off experiments to validation or A/B testing workflows.

How to use: Complete the validation checklist items to the extent possible, then enter a single hypothesis. Submit to save both the checklist status and a hypothesis record. Return to add more hypotheses or to update status as tests progress.

Give this EDA session a short name so related submissions can be tracked (e.g., 'Churn_EDA_2026-08-26').
Person completing this checklist.
Have you confirmed the origin, extraction logic, and last-refresh timestamps for each primary source?
Is there documentation describing fields, types, units and known transforms?
Have you checked for missingness by variable, by segment, and over time?
Have you inspected variable ranges, flagged impossible values, and investigated extreme points?
Is the sample representative of the target population? Have you identified likely biases?
Are the exact preprocessing steps (joins, filters, imputation, feature creation) recorded so others can reproduce the analysis?
Were random seeds, library versions and the analysis script or notebook version noted?
Have you noted where multiple testing might inflate false positives or where selective reporting could occur?
Are there clear follow-up checks (e.g., alternate definitions, holdout splits, stratified tests) you will run to validate findings?
Record missing documentation, data quality concerns, or steps needed to make this observation testable.
A short, testable hypothesis statement (e.g., 'Users who see promo A have 8% higher conversion than control').
Describe the observation, why you think it might be true, and the expected direction or effect size if relevant.
Summarize the charts, aggregates or patterns that produced this observation (include file names, notebook cells, or figure references).
How will you test this hypothesis? (e.g., A/B test, holdout validation, difference-in-differences, predictive model evaluation). Include metrics, success criteria, and sample size notes if known.
How important is this hypothesis to business outcomes or risk reduction?
Person responsible for converting the hypothesis into an experiment or validation activity.
Optional — date or sprint when this should be validated (YYYY-MM-DD or sprint name).
Current status of this hypothesis record.
Links to notebooks, dashboards, SQL queries, data slices, or tickets that help reproduce or act on this hypothesis.
You can explore this tool now. Sign in or create an account to save your responses and return to them later.
Sign In to Save Create an Account
Make this tool part of your work

Save a personal copy, bring it to your team, or tailor the questions and workflow to fit what you are hungry to improve.

Member customization and team collaboration are coming soon.
Discuss Ask the Librarian

Discussion

Comments and conversation will live here.