The Gold Standard of Archaeology | Reference-Material Aliquot Homogeneity, Minimum Sample Mass and Depletion Planning

1. Read homogeneity as a scale-dependent claim

Back to contents

A material can be acceptably homogeneous at one test-portion mass and heterogeneous below it. The certificate or supporting report defines the measurand, preparation and minimum mass for which uncertainty is supported. Do not convert bottle-level uniformity into particle-level uniformity without evidence.

First-principles check: Name the decision, the people, material or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred result. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Extract every mass, mixing and preparation condition from a certificate. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: The method consumes less than the certificate minimum. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Certificate version, measurand, unit, minimum mass, preparation and intended use. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

2. Distinguish bottles, aliquots and replicate readings

Back to contents

Between-bottle variation, within-bottle variation, subsampling variation and instrumental repeatability are different components. Repeated readings of one digest estimate the instrument, not whether two scoops from the bottle are comparable. Design evidence at the level where the material is physically divided.

First-principles check: Name the decision, the people, material or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred result. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Map variance components for powder split into bottles and test portions. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: The laboratory makes three injections from one aliquot. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Sampling hierarchy, independent units, preparation, replicates, variance and conclusion. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

3. Control mixing, segregation and moisture

Back to contents

Fine particles, density differences, electrostatic effects and transport can segregate material. Opening can change moisture, while aggressive mixing can damage some matrices. Follow certificate instructions and validate any local aliquoting process. Record settling time and tool cleaning so the preparation itself does not create gradients.

First-principles check: Name the decision, the people, material or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred result. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Write a bottle-opening and mixing protocol for sediment powder. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: The material clumps after humidity exposure. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Environment, mixing method, duration, tools, observations and deviations. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

4. Set a defensible minimum sample mass

Back to contents

Use producer evidence where available. When local work must investigate a smaller mass, sample independent positions or aliquots and estimate the additional heterogeneity contribution. A low relative standard deviation can still be misleading if the method noise masks material variation. State the permitted mass by measurand and uncertainty goal.

First-principles check: Name the decision, the people, material or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred result. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Compare 500, 200, 100 and 50 milligram portions across bottles. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: A rare analyte appears in discrete mineral grains. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Mass levels, positions, bottles, methods, variance model, limits and decision. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

5. Plan aliquots as controlled descendants

Back to contents

Every secondary vial needs identity, parent bottle, mass transferred, date, operator, container, storage and permitted use. Aliquoting can reduce contamination and repeated opening but adds transfer loss and another homogeneity assumption. Do not pool leftovers from different preparations as if they were certified material.

First-principles check: Name the decision, the people, material or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred result. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Design an aliquot register and label for twenty vials. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: One vial is partially spilled and topped up from another. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Parent-child links, quantities, containers, losses, status and reconciliation. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

6. Worked example: tiny portions preserve stock but lose representation

Back to contents

A laboratory has eight grams of soil reference material and expects forty future runs. Its method could digest fifty milligrams, but the certificate supports a minimum of two hundred milligrams. A pilot finds much larger between-portion variance at fifty milligrams for lead. The lab keeps two-hundred-milligram controls for release decisions, uses smaller portions only for training and reserves a sealed gram for a future lot bridge.

First-principles check: Name the decision, the people, material or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred result. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Calculate the run budget and classify every intended use. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: A new instrument reduces detection limits but not particle heterogeneity. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Inventory, mass rule, pilot evidence, use classes, reserve and review. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

7. Build a depletion forecast before scarcity

Back to contents

Forecast use by control frequency, methods, repeats, investigations, validation and training. Include transfer and failed-preparation losses. Set reorder and bridge triggers in mass and time, not when the last bottle is opened. Protect a reserve proportionate to replacement uncertainty and consequences.

First-principles check: Name the decision, the people, material or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred result. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Forecast three years under expected and high-investigation scenarios. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: The supplier announces discontinuation. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Starting mass, planned uses, losses, scenarios, thresholds, reserve and owner. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

8. Allocate material by decision importance

Back to contents

Use certified material where traceability or high-consequence release requires it, and qualified in-house controls where they are fit for trend monitoring. Do not ration by quietly reducing mass below support. Document prioritisation and ensure emergency investigations do not consume the bridge reserve without named approval.

First-principles check: Name the decision, the people, material or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred result. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Rank calibration, validation, daily control, training and research uses. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: A unique archaeological sample requires an urgent repeat. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Use classes, consequence, alternatives, allocation rule, exceptions and approvals. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

9. Bridge replacement lots while both are available

Back to contents

Test old and new lots across relevant measurands, methods, days and concentration ranges using independent preparations. Preserve certificate differences and expected uncertainty. The goal is to quantify comparability and update control limits, not force identical means. Overlap consumes old stock, so budget it deliberately.

First-principles check: Name the decision, the people, material or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred result. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Design a two-lot bridge with four days and two instruments. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: Only three old-lot portions remain. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Lots, design, mass, methods, acceptance, uncertainty, results and transition. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

10. Propagate sub-sampling uncertainty into results

Back to contents

When aliquot heterogeneity contributes materially, include it in control interpretation and reported uncertainty where applicable. A stable instrument cannot remove sampling variance. Investigate outliers for preparation and heterogeneity before recalibration; do not average away evidence that the material is unrepresentative at the chosen mass.

First-principles check: Name the decision, the people, material or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred result. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Build an uncertainty budget separating weighing, preparation and material variation. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: The uncertainty is use-specific across analytes. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Components, data, model, scope, sensitivity and reported statement. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

11. Audit physical and digital inventory together

Back to contents

Reconcile bottle mass, aliquot mass, uses, spills, disposals and reserves. Check that software selects the right lot, certificate and minimum mass. Repeated unexplained losses, undocumented aliquots or negative inventory are signals of broken custody and can invalidate later bridging claims.

First-principles check: Name the decision, the people, material or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred result. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Run a blind inventory reconciliation for one reference material. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: The ledger shows more remaining mass than the bottle weighs. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Physical count, ledger, certificates, uses, discrepancies, impact and correction. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

12. The gold standard spends reference material without spending credibility

Back to contents

A dwindling bottle creates pressure, but scarcity does not change the scale at which certification applies. Strong laboratories respect minimum mass, distinguish sampling from reading replication, control aliquots, forecast depletion, reserve bridge material and carry heterogeneity uncertainty into decisions.

First-principles check: Name the decision, the people, material or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred result. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Review one material from receipt to final bridged lot. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: Repeat for a limited archaeological in-house reference with no producer certificate. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Certificate, sampling evidence, aliquots, inventory, forecasts, allocation, bridge and audit. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.