From Observation to Evidence: Method

PROTOCOL / 10.0

A stronger result begins with a traceable method.

Planning, calculation, new observation, repetition and review have distinct supporting evidence. This framework does not replace assessment by actual scientific leads.

Freeze the question before the result

Record target, time system, data selection, allowed model family and comparison criterion. Participants may choose their own model settings within allowed bounds; settings are retained with the result. Opening a new record does not make a previously seen result unseen.

A schedule is a scenario

A linear transit schedule is calculated from published T₀ and P. Where covariance is absent, the independence assumption used for ranking and a separate upper bound for the window stay explicit. An entered new-measurement error is not an actual measurement. Ranking does not measure a target’s full scientific importance.

Preserve the time system

BJD_TDB, BKJD and UTC are not interchangeable numbers. The defined 2454833-day offset is retained for Kepler BKJD. Approximate UTC conversion states its numerical allowance and future leap-second assumption; measurement times are never silently relabelled as UTC.

Conditional model and actual calibration

The workshop’s formal uncertainty is conditional on its fixed model family. It does not cover every atmospheric effect, stellar spot, correlated noise or missing calibration. A calibration declaration is not independent verification that dark/flat correction was applied. Raw-FITS photometry and analysis of an existing light curve are separate operations.

Controlled access and separate accounts

Owner, two participants and optional reviewer are distinct site accounts. The owner cannot reveal a comparison before both results exist. Permissions and version bindings are rechecked on every request. A participant does not receive the other participant’s private result. Human identity and scientific expertise are not automatically verified.

Publication is a separate decision

Saving, two submissions or agreement with a server calculation is not public publication. After participant consents and a distinct review decision, an administrator separately publishes an allowed summary. Review scope stays explicit; this creates neither a DOI nor academic peer-review approval. Withdrawing access also re-evaluates public-record eligibility.

Pilot outcomes to measure

Fraction of eligible datasets yielding results and reasons for rejection.
Differences between separate measurements and the scope of stated uncertainty.
Researcher time compared with the prior method on the same scope.
Fraction with explicit data/code reuse rights and review records.

No success rate or partner count is assigned to a pilot that has not been conducted.

Primary sources and scope

These links document sources and methods; they do not declare institutional partnership or scientific endorsement.