Evidence Over Enthusiasm

The operating rule

An exciting claim becomes a question before it becomes a conclusion.

TerraEvidence records what was reported, how it was measured, where it may apply, what contradicts it, and what remains unknown. Interest can choose the next question; it cannot choose the answer.

Five different statements

Label the claim before evaluating it.

Fact

A directly verifiable detail, such as the wording, value, unit, method, date, location, organism, or condition stated in a source.

Observation

Something seen or measured in a defined setting. An observation may be accurate without establishing its cause or generality.

Hypothesis

A proposed explanation or prediction that remains open to testing, competing explanations, revision, and rejection.

Estimate

A calculated or inferred quantity whose assumptions, inputs, range, and uncertainty must remain attached.

Recommendation

An action judgment that may require evidence beyond the underlying observation, including local conditions, benefits, costs, safety, regulation, and qualified professional guidance.

From discovery to publication

The TerraEvidence review path

1Capture

Preserve the original wording, source, date, and link without silently strengthening the claim.

2Classify

Mark the material as fact, observation, hypothesis, estimate, recommendation, or unresolved combination.

3Check context

Record place, country, climate, soil or medium, organism, treatment, method, comparison, timing, and measurement conditions when available.

4Test support

Look for independent evidence, appropriate controls, comparable methods, repeat findings, limitations, null results, and contradictions.

5Assign status

Use a visible label that reflects the review completed—not the importance or excitement of the idea.

6Publish selectively

Release only material that is useful at its current evidence status. Keep raw recovery separate from public conclusions.

7Invite correction

Make sources and limits visible enough that a reader can identify an error, missing condition, or conflicting result.

8Revise openly

Update the conclusion or status when better evidence arrives. Preserve meaningful contradictions instead of erasing them.

Evidence status

The label describes review, not truth.

A status is a record of what has been checked so far. It is not a guarantee that a claim will never change.

Source record

Preserved

The item has an identifiable source and enough original context to recover what was actually said.

Developing

Being evaluated

The idea may be useful, but sourcing, method comparison, contradiction review, or scientific interpretation is incomplete.

Reviewed

Checked for stated scope

The item has been checked for wording, source traceability, method context, limits, and fit with the conclusion presented. Reviewed does not mean universally proven.

Minimum evidence record

Context that should travel with a claim

Not every source reports every field. Missing information remains marked as missing rather than being guessed.

□ Original claim, result, or observation □ Source, author or institution, date, and stable link □ Country, region, climate, and relevant setting □ Soil type, growing medium, crop, organism, or system □ Treatment, comparison, controls, replication, and sample size when reported □ Measurement method, units, timing, and reporting basis □ Limitations, missing context, uncertainty, and alternative explanations □ Contradictory, null, or less favorable findings □ Evidence status and date of the latest review

Contradictions

Disagreement is information.

Conflicting results can reveal hidden differences in soil, climate, organisms, doses, timing, methods, endpoints, or experimental design. TerraEvidence does not force unlike findings into a false average or select only the result that supports a preferred story.

Keep both results visible

Record the scope and evidence status of each result before trying to reconcile them.

Check comparability

Ask whether the studies measured the same property with compatible methods, units, treatments, and conditions.

Look for effect modifiers

Soil texture, pH, climate, crop, microbial community, dose, timing, and management history may change the response.

Allow an unresolved answer

“The evidence conflicts” is more useful than a confident conclusion built by hiding uncertainty.

Before repeating a product claim

Questions enthusiasm tends to skip

□ What exact outcome was measured—not merely described? □ Was there an appropriate comparison or control? □ Were the rate, formulation, timing, conditions, and duration reported? □ Was the evidence independent of the seller or inventor? □ Were null results, harms, tradeoffs, and uncertainties reported? □ Does the evidence match the soil, crop, climate, scale, and intended use being discussed? □ Is the conclusion stronger than the study design or measurement can support? □ Have education and evidence been kept separate from affiliate or commercial interests?

Sources and further reading

Rigor requires enough detail to inspect and repeat the work.

National Academies — Reproducibility and Replicability in Science

A cross-disciplinary consensus report examining definitions, sources of non-reproducibility, transparency, and practices that strengthen research.

Read the consensus report

USDA — Scientific Research Information Quality Guidelines

Describes quality criteria used when USDA agencies develop, review, and disseminate scientific research information.

Read the USDA guidelines

NIH — Enhancing Reproducibility through Rigor and Transparency

Collects policies and resources focused on rigorous methods, transparent reporting, reproducibility, and scientific merit.

Review the NIH resources

Source review: August 2026. These sources inform the general method; TerraEvidence remains responsible for applying and labeling its own review decisions.