What might be true?
A testable proposition states how the selected issues may interact and what would make the combination more than shared language or taxonomy.
Intersectionality Beta · research-thesis library · July 24, 2026
Research Theses & Abstracts
A searchable thesis and abstract is available for every complete k=2 comparison and every cohesive k=3, k=4, and k=5 candidate in the July 24 analytical package.
238,816 thesis records
200 workbook-authored themes
Deterministic static projection
0 verified conclusions
Research boundary. A thesis is a bounded proposition to investigate—not a finding. Abstracts summarize the metadata basis and the work still required to establish mechanism, affected population, stakeholder pathways, jurisdiction, timeframe, source fit, counterevidence, and limitations.
Complete analytical scope
From comparison record to research proposition
Every record receives a reproducible thesis and abstract. The top 200 additionally use the workbook-authored editorial theme, reporting thesis, evidence plan, impact proposition, and limitations.
- k=2 complete pairs
- 30,135 Includes candidate, netted, and low-closeness comparator cases
- k=3 candidates
- 11,289 Every internal pair exceeds 15% closeness
- k=4 candidates
- 45,851 Six constituent pair links per record
- k=5 candidates
- 151,541 Ten constituent pair links per record
- Featured themes
- 200 72 k=2 · 55 k=3 · 45 k=4 · 28 k=5
What must be studied?
The abstract records the quantitative basis, weakest-link control, scope, and the evidence questions needed before any conclusion.
Research required
Generated text never promotes a score into causation, stakeholder impact, legal equivalence, or an evidence-verified intersection.
Workbook-authored editorial opportunities
Featured research theses
The top 200 layer prioritizes public-interest reporting potential while retaining the same metadata-candidate limitations.
Loading featured theses…
Full k=2–k=5 library
Browse every thesis record
Threshold-aware static data keeps the default view light. Lowering the minimum score deliberately loads the additional records needed for that view.
k=2 thesis view
Loading research theses…
The selected static data is loading.
Loading the selected thesis records…
Method and provenance
How the thesis library is produced
The page combines workbook-authored editorial theses with a deterministic, non-AI template projection for the remaining analytical records.
Generic projection
Stable issue labels, category breadth, score, routing state, and weakest-link data are converted into a bounded thesis and abstract. No OpenAI or external provider is called by the page.
Featured override
For 200 matching candidate IDs, the site uses the uploaded workbook’s Candidate Theme and Reporting Thesis and adds its evidence plan, public-impact proposition, ranking rationale, and limitations.
Required review
Every thesis still requires a named mechanism, compatible population, jurisdiction and period, exact sources, stakeholder-specific pathways, counterevidence, and a governed disposition.
Exact files and analytical boundaries
- Research-thesis manifest
- Open manifest
- Featured layer
- Open top-200 data
- Complete k=2 index
- Open k=2 shard index
- Higher-order records
- Reuses the validated k=3, k=4, and threshold-sharded k=5 files already powering Intersectionality Beta.
- Stakeholder boundary
- The uploaded match workbooks do not enumerate why each stakeholder is affected. The abstracts therefore identify stakeholder-pathway analysis as required future work rather than inventing it.
- Demographic boundary
- No demographic distribution is inferred from issue labels or match scores.