What Are We Actually Testing?
Responsible innovation does not begin by assuming the answer is software. The goal of discovery is to learn what kind of thing this actually needs to become.
Hypotheses
Historical state voter-file records contain useful signals that are lost when we only examine the current record.
TO TESTWe can reliably reconstruct individual voter histories from recurring snapshots.
TO TESTCertain record changes predict future voter-registration instability.
TO TESTOrganizations can identify people experiencing instability quickly enough to intervene.
TO TESTOutreach can improve registration recovery or stability.
TO TESTSome interventions work better than others.
TO TESTRegistration assistance can become a meaningful entry point into longer-term organizing relationships.
TO TESTThe infrastructure can be replicated beyond one organization or one state.
TO TESTStability Signals — Candidate Signals
We do not yet know which changes actually predict future registration problems. Discovering that is part of the research. Turn candidate signals on and off to see how the flagged population changes.
2,002 people
would be flagged using these demonstration rules. Defining "instability" is itself a research problem.
What Should Trigger an Alert?
What Could We Learn?
Click a question to see the data, analysis, experiment, and decision it could support.
What If It Doesn't Work?
Those are useful findings.
An Innovation Lab exists to reduce uncertainty. A project can create value by discovering that an assumption is wrong before an organization invests heavily in scaling it.
All hypotheses, signals, and rules are candidate concepts for research — not validated models.