Evidence quality
Turn scattered observations into a reproducible record with explicit uncertainty and professional review thresholds.
Single reading versus trend
Separate a momentary value from a repeated pattern with known timing, conditions and expected variability.
Affected area versus control
Use a nearby unaffected comparison to strengthen interpretation without assuming both assemblies are identical.
Before-and-after evidence quality
Define the baseline, single intervention and predicted response before judging an improvement.
Weather context quality
Preserve exact rain, wind, temperature and humidity timing so event-linked symptoms can be tested.
Occupancy context quality
Record cooking, bathing, sleeping, doors and ventilation use that can materially change observations.
Instrument limitation statement
State what the instrument measures, stated accuracy, known interferences and what it cannot prove.
Measurement placement quality
Document height, surface, distance, orientation and stabilization so another person can repeat the observation.
Timestamp and clock quality
Keep synchronized clocks and original timestamps across photos, readings, weather and interventions.
Photographic evidence quality
Combine overview, detail, scale and unchanged originals while avoiding edits that replace the source file.
Evidence preservation quality
Preserve originals, filenames and intervention chronology without overstating legal chain-of-custody claims.
Repeatability quality
Repeat under comparable conditions and distinguish consistent response from sensor or observer noise.
Confounder control
Change one variable at a time and record simultaneous weather, occupancy or equipment changes.
Negative evidence
Record where and when the symptom did not occur because absence under defined conditions can separate pathways.
Uncertainty labeling
Use explicit confidence language and unresolved alternatives instead of converting signals into certainty.
Source and transport separation
Keep source, transport route, affected material and visible symptom as separate claims.
Hypothesis-led evidence
Write competing explanations and predictions before collecting more data to reduce confirmation bias.
Repair verification quality
Define recurrence windows, trigger conditions and control areas before declaring an intervention successful.
Professional handoff quality
Deliver a concise chronology, evidence table, limitations, actions and questions that a professional can audit.
