Skip to content

Q001 — ACH Matrix

Evidence vs. Hypotheses Matrix

Evidence H1 Complete prompts exist H2 No such prompts exist H3 Partial implementations
SRC01-E01 — 58 techniques, no research framework category -- ++ +
SRC02-E01 — 3 SATs via LLM, code-based - - ++
SRC03-E01 — PRISMA-trAIce reporting standard N/A + N/A
SRC04-E01 — Zero analytical framework in Perplexity -- ++ +
SRC05-E01 — Complete pipeline in code - N/A ++
SRC06-E01 — Autonomous research system - N/A ++

Legend: ++ Strongly supports, + Supports, - Contradicts, -- Strongly contradicts, N/A Not applicable

Diagnosticity Analysis

  • Most diagnostic evidence: SRC01-E01 (comprehensive taxonomy gap) and SRC04-E01 (leading tool lacks framework) are most diagnostic because they strongly discriminate between H1 and H2/H3.
  • Least diagnostic evidence: SRC03-E01 has limited diagnosticity as it addresses a different aspect (reporting standards vs. operational prompts).
  • Evidence cluster: SRC05-E01 and SRC06-E01 form a cluster showing code-based alternatives exist, which is diagnostic between H2 (too absolute) and H3 (nuanced).

Outcome

H3 (Partial implementations exist) has the most consistent support across all evidence items, with no contradicting evidence. H1 (Complete prompts exist) has the most contradicting evidence. H2 (No such prompts exist) is contradicted by evidence of implementations that go beyond narrow tasks, though they are partial and code-based rather than prompt-based.