Companion to an archived preprint
The question
At the same false-alarm allowance, do generic warning indicators identify transitions more often than no-transition controls?
The reported result
The study reports that its generic early-warning suite did not clear the statistical significance criterion across the tested modalities under the matched-false-alarm protocol. A visible trend in a selected record is a different question from reliable discrimination against controls.
Read the original publication ↗
How to read the experiment
- Declare transition and control records
- Match the false-alarm budget
- Compare detection with a permutation test
What the result does not establish
The conclusion is bounded to the tested detectors, corpora and evaluation protocol. It does not establish that every early-warning method is ineffective in every domain.
Follow the evidence
- DOI record, paper and available files
- Project publication context
- Source repository and reproduction material
- ANULUM BibTeX collection
For reproduction, use the original paper’s protocol and repository instructions, including its data and hardware requirements. The browser learning demos explain concepts; they do not reproduce these experimental studies.