Cannabis sampling and measurement uncertainty: why one number does not describe a whole batch
How sample identity, homogeneity, moisture and measurement uncertainty affect batch comparisons, with a practical review checklist.
A batch is not the laboratory sample
A laboratory first examines the material actually received. If a statement is intended to cover a larger batch, sampling must address that aim. An unusual flower or material left in an open container is not automatically representative. Define the batch, the intended conclusion and who collected the material. Representativeness is a question that arises before the instrument. Eurachem therefore considers sampling, physical preparation and analysis together as the measurement process. A precise laboratory value cannot by itself correct an unsuitable selection of material.
Document differences openly
Samples may differ in collection point, storage, product form and preparation. A useful comparison keeps conditions consistent or records differences. Note the batch identifier, date, container and known handling steps. Such records do not guarantee a perfect sample; they make differences traceable. A sampling instruction must be confirmed by the responsible laboratory for the actual purpose. CannabisLabor.ch currently provides no submission or collection instructions and accepts no products. Do not improvise a shipment based on an educational article.
Moisture changes the reference basis
A concentration in the original material and one on a dry-matter basis use different references. Drying and moisture determination therefore matter to interpretation. The NIST CannaQAP moisture comparison reports differences between drying approaches used by laboratories. For a practical review, compare the declared moisture basis and procedure as well as the percentage. A higher value cannot automatically be interpreted as a better or stronger batch. Preserve the complete report so that the basis remains attached to the figure.
What measurement uncertainty means
A result is not infinitely precise. Reported uncertainty describes dispersion associated with the result under specified assumptions. Ask whether an interval covers the analytical step alone or sampling as well, and how it should be interpreted. Slightly different figures do not automatically establish a meaningful difference between batches. Equally, overlapping intervals do not prove that products are identical. A specific statistical assessment requires full context and qualified interpretation, rather than a simple visual comparison of two percentages.
Make comparisons with clear limits
Before testing, decide whether the aim is to assess a delivery, compare a process or investigate a label statement. Agree the sampling plan and report scope accordingly. Keep original findings instead of only copying numbers into a spreadsheet. Where information is missing, ask questions before making a commercial commitment from the comparison. The same care applies to private products. These questions help preparation; they replace neither a confirmed laboratory procedure nor an assessment of health safety.
| Document | Demo value | Reference basis | Open question |
|---|---|---|---|
| Report A | 10.0% | Original material | How was the sample collected? |
| Report B | 11.0% | Dry matter | Is the moisture basis comparable? |
Questions to clarify locally
- What exactly defines the batch?
- Who collected the sample, where and when?
- Are the matrix, unit and moisture basis comparable?
- Which parts of the process does the uncertainty cover?
- Is the difference meaningful for the actual question?