| name | validity-probing |
| description | Challenge construct validity — does benchmark measure claimed capability? — 3 benchmarks, 40 papers, 30 web searches |
| used-by | benchmark-archaeology |
Validity Probing Strategy
Deep investigation of construct validity for individual benchmarks. Determines whether a benchmark actually measures the capability it claims to measure, or whether high scores can be achieved through shortcuts, artifacts, or unrelated competencies.
Purpose
Produce a construct validity assessment that identifies the gap between what a benchmark claims to measure and what it actually measures. Expose confounds, shortcuts, and alternative explanations for high performance.
Budget
| Resource | Floor | Target |
|---|
| Benchmarks probed | 2 | 3 |
| Papers read | 30 | 40 |
| Web searches | 20 | 30 |
State Ledger
<HARD-GATE>
| Metric | Current | Target | Status |
|--------|---------|--------|--------|
| Benchmarks probed | 0 | 3 | PENDING |
| Papers fetched | 0 | 40 | PENDING |
| Papers read | 0 | 30 | PENDING |
| Web searches | 0 | 30 | PENDING |
| Construct validity assessments | 0 | 3 | PENDING |
| Artifact detection runs | 0 | 3 | PENDING |
| Alternative explanation catalogs | 0 | 3 | PENDING |
| Convergent validity checks | 0 | 3 | PENDING |
</HARD-GATE>
Cannot exit until 80% of all targets met.
Available Tactics
- artifact-detection — Detect annotation artifacts and shortcuts
- evaluation-protocol-comparison — Compare how different papers implement the benchmark
Available SOPs
- construct-validity-assessment — Core validity evaluation
- metric-decomposition — Understand what the metric actually captures
- contamination-audit — Rule out data leakage as confound
- benchmark-synthesis — Produce validity report
Execution Guidance
- Target Selection: Choose 3 benchmarks where validity concerns exist (high scores but questionable real-world transfer, known shortcuts, or contested claims)
- Per-Benchmark Deep Dive:
a. Collect the original benchmark paper + all critique/analysis papers
b. Run construct-validity-assessment: map claimed capability to actual task requirements
c. Run artifact-detection tactic: probe for shortcuts and spurious correlations
d. Run metric-decomposition: identify what signals the metric actually rewards
e. Check convergent validity: do models that score high also perform well on related tasks?
f. Check discriminant validity: do models that score high fail on tasks they shouldn't if the capability were real?
- Alternative Explanation Catalog: For each benchmark, enumerate non-target explanations for high scores
- Synthesis: Produce validity verdict with evidence strength ratings
Output Format
validity_report:
benchmark_name: string
claimed_capability: string
actual_measurement: string
validity_verdict: valid|partially_valid|questionable|invalid
evidence_strength: strong|moderate|weak
confounds_identified:
- confound: string
severity: high|medium|low
evidence: string
shortcuts_found:
- shortcut: string
exploit_method: string
performance_gain: string
convergent_validity: pass|partial|fail
discriminant_validity: pass|partial|fail
recommendations: list[string]