Benchmarks / Civil liberties (SPK + LIB) · Survey

GSS (2024) — Civil liberties (SPK + LIB) batteries

https://gss.norc.org/get-the-data/ →

benchmark:gss-free-speech-2024

Reuse this data: CSV JSON headers match the published fixture; distributions JSON-encoded inline

Shape mix: 4 bimodal-asymmetric 4 multimodal-symmetric

Δ 1.53 4.18 () → 2.65 () within-fixture gradient

Acceptance rubric

An SSR re-run acceptably reproduces this benchmark when it clears every check below on the same 8 paired questions the source publishes. The chip badges name the checks the round-38 → round-46 shape-mismatch finding argued are necessary but not sufficient in isolation.

CI overlap 8/8 Shape match 8/8 SSR mean inside CI · shape badge reproduces the source label.

What this benchmark discloses

Recruitment screener (verbatim)

GSS uses an address-based national probability sample of U.S. adults 18+. The 2024 cross-section ran mixed-mode (in-person + web push). No content screener — every selected adult is invited regardless of attitudes. Sampling frame and selection probability are documented in the GSS 2024 Codebook R3 (NORC). SPK/LIB items are split-ballot (per-item n ≈ 1,100–1,130, vs the wave's 3,309 total).

Sample weighting & demographics

Weighted to wtssps (NORC's recommended GSS 2024 post-stratification weight, person-level). Tolerant share is the weighted proportion answering 'allowed to speak' (SPK) or 'would not favor removing' (LIB). Dichotomous → 1-5 SSR mapping via tolerant_share × 4 + 1.

Fieldwork dates, country, language

United States, 2024 cross-section. Fieldwork March–November 2024 per the GSS 2024 Codebook R3 (page 1). Modes: in-person and web push (push-to-web for the subsample without prior in-person contact).

Original question text & translations

SPK stem (codebook verbatim): "If such a person wanted to make a speech in your community, should he/she be allowed to speak, or not?" LIB stem: "If some people in your community suggested that a book he/she wrote should be taken out of your public library, would you favor removing this book, or not?" The target category (atheist / communist / racist / anti-American Muslim clergyman) substitutes 'such a person'.

Likert / scale labelling

Dichotomous (codebook value labels, verbatim). SPK: 1 = "allowed", 2 = "not allowed" — tolerant_code = 1. LIB: 1 = "favor removing", 2 = "would not remove" — tolerant_code = 2 (reverse). Both remapped to the SSR 1-5 scale via tolerant_share × 4 + 1 so that a higher number reads as more tolerance. No 'don't know' / 'no answer' is included — those respondents drop from each variable's n.

FGI moderator guide & coding rules

Closed-ended forced-choice, no FGI. No moderator guide. Each respondent rates each category × battery cell independently in the same wave. SPK and LIB have run since 1972 / 1976 with the same stems and value labels; the 2024 wave dropped homo, militarist, and the full COL battery — only 8 of the historical 15 cells are present.

Data source & license

Public domain (NSF-funded). Required citation: Davern, Michael; Bautista, Rene; Freese, Jeremy; Herd, Pamela; and Morgan, Stephen L. General Social Survey 1972-2024. NORC ed., Chicago, 2025.

Anything not disclosed here was not available in the source dataset and is recorded as such.

SSR run — reproducibility keys

template_version: 3.0.0
llm_seed: 4212
bootstrap_seed: 8424
llm_model: deepseek-v4-flash
embedding_model: nvidia/nv-embed-v1

SSR score vs ground truth, by question

Each row holds one question. The teal mark and range are SSR. The sky mark and range are the source FGI/survey.

§ If a person who is against all churches and religion wanted to make a speech in your community, should he be allowed to speak, or not?

SSR —

Source 4.18 (4.09–4.28)

Source distribution SD 1.61 skew -1.47 2 cells bimodal-asymmetric

1.02.03.04.05.0

§ If an admitted Communist wanted to make a speech in your community, should he be allowed to speak, or not?

SSR —

Source 3.60 (3.49–3.72)

Source distribution SD 1.91 skew -0.63 2 cells bimodal-asymmetric

1.02.03.04.05.0

§ If a person who believes that Blacks are genetically inferior wanted to make a speech in your community, should he be allowed to speak, or not?

SSR —

Source 2.84 (2.72–2.96)

Source distribution SD 1.99 skew 0.16 2 cells multimodal-symmetric

1.02.03.04.05.0

§ If a Muslim clergyman who preaches hatred of the United States wanted to make a speech in your community, should he be allowed to speak, or not?

SSR —

Source 2.65 (2.53–2.76)

Source distribution SD 1.97 skew 0.36 2 cells multimodal-symmetric

1.02.03.04.05.0

§ If some people in your community suggested that a book written by an anti-religion atheist should be taken out of your public library, would you favor removing this book, or not?

SSR —

Source 4.12 (4.02–4.22)

Source distribution SD 1.66 skew -1.35 2 cells bimodal-asymmetric

1.02.03.04.05.0

§ If some people in your community suggested that a book written by an admitted Communist should be taken out of your public library, would you favor removing this book, or not?

SSR —

Source 3.78 (3.67–3.89)

Source distribution SD 1.84 skew -0.85 2 cells bimodal-asymmetric

1.02.03.04.05.0

§ If some people in your community suggested that a book written by a racist (who believes Blacks are genetically inferior) should be taken out of your public library, would you favor removing this book, or not?

SSR —

Source 3.23 (3.11–3.34)

Source distribution SD 1.99 skew -0.23 2 cells multimodal-symmetric

1.02.03.04.05.0

§ If some people in your community suggested that a book written by a Muslim clergyman who preaches hatred of the United States should be taken out of your public library, would you favor removing this book, or not?

SSR —

Source 3.09 (2.97–3.20)

Source distribution SD 2.00 skew -0.09 2 cells multimodal-symmetric

1.02.03.04.05.0

← All benchmark studies