The Observer & The Data: Exploring Worldviews Through Voter Rolls
Data is effectively dead until it is observedโand those who observe inevitably bias the output via their specific inquiries. At Marigold Insights, we treat public voter rolls not as a weapon for partisan attack, but as a mirror reflecting different values across the American political spectrum.
Select an Ideological Inquiry Lens
Filter the questions below to see how different groups approach election integrity and why every perspective adds value to overall system health.
The Verification Lens
Focuses on preserving confidence by ensuring strict compliance with physical residence mandates and eliminating stale records before ballots are mailed.
The Institutional Lens
Focuses on administrative excellence, seamless county-to-county data transfers, and resolving bureaucratic traps without punitive friction.
The Enfranchisement Lens
Focuses on protecting eligible voters from erroneous purges, ensuring equal polling place access, and safeguarding student or mobile populations.
Statistical Models Explained for Everyone
You do not need a PhD in statistics to audit your local government. Here is how our automated math models work in simple, everyday language.
1. The Bell Curve Alarm (Z-Scores)
Standard DeviationImagine finding the average number of people living in a single household across your town (usually around 2 or 3). If an address has 15 voters registered to it, the **Z-Score** measures how many "steps" away from normal that house is. Anything over 3 steps is flagged as a statistical alarm requiring human review.
2. The Tukey Fence (IQR Quantiles)
Robust OutliersSometimes a town has a massive college dorm or nursing home that skews the average for everyone else. **Interquartile Range (IQR)** lines up all addresses from smallest to largest, chops off the bottom 25% and top 25%, and builds a mathematical "fence" around typical homes. This prevents false alarms on normal apartments.
3. Lightning Strike Odds (Poisson Math)
Rare Event ProbabilityIf your rural county typically sees 3 new voter registrations on a normal Tuesday, what are the exact odds that 450 registrations arrive on a single afternoon? **Poisson Math** calculates rare event probabilities, helping auditors distinguish between legitimate registration drives and computer file upload glitches.
4. The Shift Detector (CUSUM)
Time-Series ChangeInstead of looking at just one day, **CUSUM** tracks cumulative trends over months. If a county election clerk suddenly stops processing address updates or alters their routine maintenance procedures, this model pinpoints the exact week the historical trend line broke.
Ready to Observe Your Local Data?
Choose your lens, run pre-configured audits, and contribute to non-partisan election transparency today.