Revisiting the sampling, sample preparation, and analytical variability associated with testing wheat for deoxynivalenol
Bibliographic record
Abstract
Fifteen lots of wheat were sampled to characterise the total variance and distribution among sample test results associated with measuring deoxynivalenol (DON) in bulk wheat lots. An unbalanced nested experimental design based on past research was used to determine contributions to the total variance from sampling, sample preparation, and analysis. The wheat lots used in the study contained average DON concentrations that ranged from 0.17 to 24.5 mg/kg. Sampling was determined to be the largest contributor to the total variance of measuring DON at low mg/kg concentrations, which are relevant to existing maximum levels. With the experimental design parameters of 1 kg laboratory samples, sub-division of whole and ground grain using rotary sample division, sample comminution using a commercial-grade coffee grinder, extraction of 100 g test portions, and making one measurement of DON in the test portion by gas chromatography-mass spectrometry, the total variance of DON measurement at 2 mg/kg was 0.046 mg2/kg2 (coefficient of variation=10.7%). At this concentration, sampling contributed 67% to the total variance, followed by sample preparation (18%) and analysis (15%). The DON distribution among sample test results was accurately described by the normal distribution. The mathematical model of variance was used with the normal distribution of DON measurement results to construct operating characteristics curves to model the likelihood of mischaracterising a wheat lot as (non) compliant with a certain decision limit. With realistic laboratory sample and test portion sizes, as well as a practicable decision limit of 1.5 mg/kg, the estimated probability of mischaracterising a wheat lot containing 2 mg/kg DON as less than this concentration was reduced to 1%.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.018 | 0.031 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".