High annotation accuracy does not by itself guarantee valid downstream inference; treating imperfect predicted labels as observed variables can bias estimates and invalidate confidence intervals, while design-based validation with a probability-sampled human-labelled subset can restore valid inference under stated assumptions.

SocioLens project,

Claim LD07-HQ-C01 (v1.0)

Provenance and status

Claim type: Fact
Publication status: Eligible for publication
Verification state: Checked against the primary source. Who checked this claim is not recorded for it; claim checks so far were AI-led, and no person has checked claims against their sources.
Does not license: Claim that human labels are construct-valid; claim that every task requires DSL; skipping human validation
Limitations: Requires corpus up front, probability-sampled gold labels and design assumptions; gold assumed high quality
Source (as cited): Egami, Hinck, Stewart & Wei, “Using Imperfect Surrogates for Downstream Inference: Design-based Supervised Learning for Social Science Applications of Large Language Models,” NeurIPS 2023.
Source type: Peer-reviewed conference paper (recorded as “Peer-reviewed conference methods paper”)
Institutional position: Independent academic
Maturity of the evidence: Rapidly maturing evidence
Recorded date: 25 September 2026
Currentness / correction: This is the current version. No correction or concern is recorded on its sources.

Sources this version cites

Egami, Hinck, Stewart & Wei, “Using Imperfect Surrogates for Downstream Inference: Design-based Supervised Learning for Social Science Applications of Large Language Models,” NeurIPS 2023 (Peer-reviewed conference paper)
Read the original

Version

This page shows version v1.0. Its record never changes, even when newer evidence arrives; the site may add a note to the page or withhold a figure not cleared for public use.

This pinned version is also the current version.

Versions: v1.0 · current

Raw record: every recorded field, with its coded values

All 25 canonical Claim fields, verbatim, with the coded values derived from them. The plain-language labels above are read from these.

Claim type (coded): FACT (as printed: “FACT”)
Publication status (coded): ELIGIBLE (as printed: “ELIGIBLE”)
Verification state (coded): CONTENT_VERIFIED_PRIMARY
Source type (coded): PEER_REVIEWED_CONFERENCE_PAPER
Maturity (coded): RAPIDLY_MATURING

#	Canonical field	Canonical value (verbatim)	Normalized (Claim Value Vocabulary v1.0.2)
1	CLAIM ID	LD07-HQ-C01	—
2	PUBLIC-FACING CLAIM	High annotation accuracy does not by itself guarantee valid downstream inference; treating imperfect predicted labels as observed variables can bias estimates and invalidate confidence intervals, while design-based validation with a probability-sampled human-labelled subset can restore valid inference under stated assumptions.	—
3	CLAIM TYPE	FACT	base_types: FACT
4	RESEARCH STRAND	A — Annotation/coding/classification	—
5	MATURITY	RAPIDLY MATURING	states: RAPIDLY_MATURING
6	SOURCE	Egami, Hinck, Stewart & Wei, “Using Imperfect Surrogates for Downstream Inference: Design-based Supervised Learning for Social Science Applications of Large Language Models,” NeurIPS 2023.	source_ids: SRC-LD07-EGAMI-DOWNSTREAM-INFERENCE-2023
7	SOURCE TYPE	Peer-reviewed conference methods paper	classes: PEER_REVIEWED_CONFERENCE_PAPER
8	INSTITUTIONAL POSITION	Independent academic	—
9	SCOPE	Downstream inference using predicted text labels across CSS tasks	—
10	DESIGN	Design-based statistical methods + benchmark evaluation	—
11	SAMPLE / DATASET	18 computational social-science tasks	—
12	MODEL / TOOL / VERSION	Surrogate classifiers / GPT-era predicted labels; claim is method-general	—
13	COMPARATOR	High-quality gold labels	—
14	MAIN RESULT	Direct surrogate use yields substantial bias/invalid CIs even at 80–90% accuracy; DSL produces low bias and approximately nominal coverage under assumptions	—
15	INFERENTIAL TARGET	Coefficients and confidence intervals using predicted variables	—
16	DOES NOT LICENSE	Claim that human labels are construct-valid; claim that every task requires DSL; skipping human validation	—
17	VERIFICATION STATE	CLEARED	state: CONTENT_VERIFIED_PRIMARY · verified_by: UNKNOWN
18	LIMITATIONS	Requires corpus up front, probability-sampled gold labels and design assumptions; gold assumed high quality	—
19	CONTRADICTORY / SUPPORTING SOURCES	Supported conceptually by Baumann; Yang; Moore	—
20	GEOGRAPHY / LANGUAGE	Primarily English CSS benchmarks	—
21	DISTRIBUTIONAL LEDGER	Subgroup validity depends on gold-sample design; not independently established here	—
22	R01 / FIELD-MAP LINKS	T1, D9, R-1	—
23	WHAT WOULD CHANGE THIS CLAIM?	Demonstration that direct predicted-label inference remains calibrated across broad tasks without correction	—
24	PUBLICATION STATUS	ELIGIBLE	state: ELIGIBLE
25	RECORDED DATE	2026-09-25	—

Field 22 references resolved against Field Map v0.2: T1 (TENSION); D9 (DOMAIN); R-1 (LENS)

The sources as recorded, with their record IDs and coded source types:

Pinned Sources (field 6, version-specific)

“Egami, Hinck, Stewart & Wei, “Using Imperfect Surrogates for Downstream Inference: Design-based Supervised Learning for Social Science Applications of Large Language Models,” NeurIPS 2023.” → Egami et al. (2023), “Using Imperfect Surrogates for…” (SRC-LD07-EGAMI-DOWNSTREAM-INFERENCE-2023--v1.0) — Egami, Hinck, Stewart & Wei, “Using Imperfect Surrogates for Downstream Inference: Design-based Supervised Learning for Social Science Applications of Large Language Models,” NeurIPS 2023; class PEER_REVIEWED_CONFERENCE_PAPER (explicit public provenance pin)

Cited by Living Debate versions

AI in the research apparatus (v0.1) LD-07
AI in the research apparatus (v0.2) LD-07

Cite this page

SocioLens project. (2026, September 25). High annotation accuracy does not by itself guarantee valid downstream inference; treating imperfect predicted labels as observed variables can bias estimates and invalidate confidence intervals, while design-based validation with a probability-sampled human-labelled subset can restore valid inference under stated assumptions [Claim LD07-HQ-C01, version v1.0]. SocioLens. https://sociolens-web.pages.dev/claims/ld07-hq-c01/v1.0/

@misc{sociolens_ld07_hq_c01_v1_0,
  author = {{SocioLens project}},
  title = {High annotation accuracy does not by itself guarantee valid downstream inference; treating imperfect predicted labels as observed variables can bias estimates and invalidate confidence intervals, while design-based validation with a probability-sampled human-labelled subset can restore valid inference under stated assumptions.},
  year = {2026},
  date = {2026-09-25},
  howpublished = {\url{https://sociolens-web.pages.dev/claims/ld07-hq-c01/v1.0/}},
  note = {Claim LD07-HQ-C01, version v1.0}
}

The link points to this version, whose record never changes. For a finding, cite the original study as well: SocioLens summarizes research and does not replace it.

Research-of-record provenance

Materialized by sociolens-materializer 0.6.0 at 2026-09-28T20:00:00.000Z; normalization contract SocioLens — Claim Value Vocabulary v1.0.2 — Pre-Spike Materialization Contract (v1.0.2). Content seal SHA-256 67506bc7b194154a61f3b79024bbcd1b8e1eb4e40ff0da52b4b769eb5ee11438.

Frozen Drive-derived inputs

R03 — LD-07 Canonical Claim Ledger v1.0 — LOCKED (v1.0) — Drive ID 1xVz_LrQAA8EA1tPhVWIrDgOUU3Zidlg9; revision: NOT_EXPOSED_BY_DRIVE_TEXT_EXPORT; Drive modified 2026-09-29T04:31:14.120Z; block “ENTRY 01 — LD07-HQ-C01” SHA-256 9e7ed2974a51b2cda1aa8188a97e45c44666f78ee75b2ccb6a7c44a51e2a903c; SocioLens — Programme Status & Authority Register v0.6 status “CURRENT.”; frozen fixture fixtures/frozen/r03-ld07-claim-ledger-v1.0-canonical.

Builder decisions (not Drive data)

VERSION = v1.0 — basis: Claim version taken from the canonical ledger artifact version (R03 — LD-07 Canonical Claim Ledger v1.0); Implementation Architecture Selection §5
CURRENT_POINTER = true — basis: SocioLens — Programme Status & Authority Register v0.6 status for "R03 — LD-07 Canonical Claim Ledger v1.0 — LOCKED": "CURRENT."
FIELD_6_SOURCE_REFERENCE = SRC-LD07-EGAMI-DOWNSTREAM-INFERENCE-2023--v1.0 — basis: field 6 of ENTRY 01 — LD07-HQ-C01 cites "Egami, Hinck, Stewart & Wei, “Using Imperfect Surrogates for Downstream Inference: Design-based Supervised Learning for Social Science Applications of Large Language Models,” NeurIPS 2023."; the frozen field-6 segment of this ClaimVersion's ledger block; Drive-derived identity authority (Issue #39 Source inventory): Batch 01 §1 "Primary source:" + publication status "PEER-REVIEWED CONFERENCE PAPER"; ledger C01 fields 6/7/8; KEY EVIDENCE.
SOURCE_PROVENANCE_PIN = SRC-LD07-EGAMI-DOWNSTREAM-INFERENCE-2023--v1.0 — basis: Issue #5 proof rendering (Architecture Proof Protocol v0.1 AP-08 / Amendment v1.2 §6): explicit version-pinned public provenance of this ClaimVersion — the pinned field-6 Source record "Egami, Hinck, Stewart & Wei, “Using Imperfect Surrogates for Downstream Inference: Design-based Supervised Learning for Social Science Applications of Large Language Models,” NeurIPS 2023."; Drive-derived identity authority (Issue #39 Source inventory): Batch 01 §1 "Primary source:" + publication status "PEER-REVIEWED CONFERENCE PAPER"; ledger C01 fields 6/7/8; KEY EVIDENCE.