7 / 43 studies state whether the plant was fresh or dried
How often each preparation detail is stated
Solid = stated · hatched = not stated (unknown) · whisker = 95% interval. Hover, focus or tap a bar for n.
- FiltrationFiltration5 of 43 studies state it (12%)95% interval: 5%–24%5/43 · 12%
Provisional: not independently validatedChecked only by an automated or unverified check. Useful for finding problems; never a validation.A second coder's reading of 40 studies stands in for the truth; that reading is itself unverified. Not a validation.Sensitivity check (provisional): not computed: filtration was not part of the provisional check
- Fresh or driedFresh or dried7 of 43 studies state it (16%)95% interval: 8%–30%7/43 · 16%
Provisional: not independently validatedChecked only by an automated or unverified check. Useful for finding problems; never a validation.A second coder's reading of 40 studies stands in for the truth; that reading is itself unverified. Not a validation.Sensitivity check (provisional): 19% (95% interval 8%–30%) if the extractor's error rates from the provisional check applied. On the 40 checked studies: 11 stated by the extractor, 11 by the second coder.
- HeatHeat11 of 43 studies state it (26%)95% interval: 15%–40%11/43 · 26%
Provisional: not independently validatedChecked only by an automated or unverified check. Useful for finding problems; never a validation.A second coder's reading of 40 studies stands in for the truth; that reading is itself unverified. Not a validation.Sensitivity check (provisional): 24% (95% interval 14%–35%) if the extractor's error rates from the provisional check applied. On the 40 checked studies: 12 stated by the extractor, 12 by the second coder.
- Plant partPlant part21 of 43 studies state it (49%)95% interval: 35%–63%21/43 · 49%
Provisional: not independently validatedChecked only by an automated or unverified check. Useful for finding problems; never a validation.A second coder's reading of 40 studies stands in for the truth; that reading is itself unverified. Not a validation.Sensitivity check (provisional): 49% (95% interval 33%–66%) if the extractor's error rates from the provisional check applied. On the 40 checked studies: 16 stated by the extractor, 20 by the second coder.
- Solvent(inflated)Solvent37 of 43 studies state it (86%)95% interval: 73%–93%Inflated by construction: a study counts as having prepared the plant partly because a solvent was stated, so this share is higher than it would be among all studies that prepared the plant.37/43 · 86%
Provisional: not independently validatedChecked only by an automated or unverified check. Useful for finding problems; never a validation.A second coder's reading of 40 studies stands in for the truth; that reading is itself unverified. Not a validation.Sensitivity check (provisional): 79% (95% interval 62%–93%) if the extractor's error rates from the provisional check applied. On the 40 checked studies: 20 stated by the extractor, 22 by the second coder.
Solvent share is inflated by construction. A study counts as having prepared the plant partly because a solvent was stated, so this share is higher than it would be among all studies that prepared the plant.
None of the five stated: 2 of 43 · 5% (95% interval 1%–15%)
View as table
| Detail | Stated | Studies (n) | Proportion | 95% interval | Sensitivity check (provisional) |
|---|---|---|---|---|---|
| Filtration | 5 | 43 | 12% | 5%–24% | not computed: filtration was not part of the provisional check |
| Fresh or dried | 7 | 43 | 16% | 8%–30% | 19% (95% interval 8%–30%) if the extractor's error rates from the provisional check applied. On the 40 checked studies: 11 stated by the extractor, 11 by the second coder. |
| Heat | 11 | 43 | 26% | 15%–40% | 24% (95% interval 14%–35%) if the extractor's error rates from the provisional check applied. On the 40 checked studies: 12 stated by the extractor, 12 by the second coder. |
| Plant part | 21 | 43 | 49% | 35%–63% | 49% (95% interval 33%–66%) if the extractor's error rates from the provisional check applied. On the 40 checked studies: 16 stated by the extractor, 20 by the second coder. |
| Solvent (inflated by construction) | 37 | 43 | 86% | 73%–93% | 79% (95% interval 62%–93%) if the extractor's error rates from the provisional check applied. On the 40 checked studies: 20 stated by the extractor, 22 by the second coder. |
| None of the five stated | 2 | 43 | 5% | 1%–15% | Not computed |
Prepared set: the 43 studies that prepared the plant material (audit v2 eligibility). "Stated" means the rule-based extractor coded a value (heuristic; checked only against a second coding of a study sample). 95% Wilson intervals. Descriptive findings, not hypothesis tests. Source: data/interim/m5/reporting_gap.json.
Other study sets
- Every coded study retrieved for a plant–use pair (audit v1): none of the five stated131 of 303 · 43% (95% interval 38%–49%)
- Heat stated · studies coded from abstracts9 of 241 · 4% (95% interval 2%–7%)
- Heat stated · studies coded from open-access methods28 of 62 · 45% (95% interval 33%–57%)
- Compared two preparations of the same plant material (pre-registered v3 screening)6 of 257 · 2% (95% interval 1%–5%)
When they did say, did it match the recipe?
Audit v2 · pre-registered re-analysis after v1's design flaw. Small n: only 11 (plant–use pair, study) units have a known status; 32 are unknown and 4 have no modern analogue. Too few to estimate a field-wide rate. Audit v1 (as pre-registered, with its design flaw) is in Methods & data.
View as table
| Detail | Same | Different | Not stated | No modern analogue |
|---|---|---|---|---|
| Solvent | 2 | 5 | 27 | 1 |
| Heat | 0 | 2 | 33 | 0 |
| Filtration | 3 | 0 | 32 | 0 |
View as table
| Detail | Same | Different | Not stated | No modern analogue |
|---|---|---|---|---|
| Solvent | 2 | 5 | 27 | 1 |
| Heat | 0 | 2 | 33 | 0 |
| Filtration | 3 | 0 | 32 | 0 |
View as table
| Detail | Same | Different | Not stated | No modern analogue |
|---|---|---|---|---|
| Solvent | 2 | 3 | 8 | 3 |
| Heat | 0 | 0 | 16 | 0 |
| Filtration | 0 | 1 | 15 | 0 |