Laboratory FieldState replication protocol

A preregistered way to test whether measured laboratory field conditions moderate a defined biological experiment — without assuming that they explain the wider replication literature.

This protocol uses a laboratory FieldState record to describe the physical conditions of a direct, blinded comparison. The candidate moderator is the measured field condition, not the record itself.

Its goal is modest and discriminating: establish whether a documented difference in measured physical field conditions changes a pre-specified endpoint under otherwise matched experimental conditions.

1. Design a matched comparison

Use at least two independently characterised environments or exposure chambers: a reference/sham condition and an active condition. Randomise samples or animals across runs and blind endpoint assessment wherever practical.

Do not describe either arm as “EMF-free”. The comparison must report what is measured in each arm, including residual fields and uncertainty.

2. Measure the full experimental environment

Record calibrated field measurements before, during and after each biological run: relevant bands, low-frequency components, B₀ vector where relevant, orientation, time series/PSD, source configuration, temperature, humidity, vibration, light and airflow.

If shielding is used, characterise its side effects. Shielding can alter temperature stability, ventilation, acoustic environment, static fields and handling; those changes need matched controls rather than being attributed to a field difference by default.

3. Pre-specify a close endpoint

Choose an endpoint that corresponds to a registered link: for example calcium/ROS dynamics, a barrier tight-junction measure, sperm function, oocyte redox or a circadian marker. Define collection time, transformations, exclusions and the primary contrast before unblinding.

A rescue or blocker arm can test mediation but cannot substitute for a sham-controlled physical contrast. A national fertility series is not a laboratory endpoint.

4. Analyse the FieldState contrast, not an assumed scalar

Report the observed differences in the named FieldState features and their uncertainty. Test the primary endpoint against the preregistered contrast; treat secondary spectral, orientation and timing analyses as exploratory unless they were prespecified.

Repeat the protocol in an independent laboratory and, where possible, with a second instrument chain. The protocol does not predict a fixed fold-change: the magnitude is an empirical question.

5. Release an auditable package

Publish the preregistration, raw or access-controlled raw field recordings, calibration certificates, chamber logs, biological data, analysis code, randomisation/blinding record, deviations and null results.

Classify input quality as partial or measurement-ready FieldState data. Measurement-ready data still require replication and an independently interpretable endpoint before they can inform an organ-state mapping.

What either result would mean

A reproducible null result under a well-characterised FieldState contrast would constrain the proposed link for that system and protocol. A reproducible difference would motivate mechanism and dose/geometry work; it would not by itself establish a human reproductive or population-TFR effect. Both outcomes are informative and should be published.