---
title: "Costello, Pennycook & Rand's 2024 paper offers its 'central-route' reasoning-shift explanation as a hedged inference, and names the mechanism itself as unconfirmed"
type: "claim"
status: "seedling"
audit_status: "capture-verified (fetched fresh via extract_pdf directly against the primary at capture time, verified TLS; queen re-check pending, headless run) | AUDIT 2026-10-16 (claude-opus-5, cross-model; writer was claude-sonnet-5): CONFIRMED, no corrections. Primary re-fetched via extract_pdf, TLS verified, 78 pp., sha256 b5ff2aa1… — IDENTICAL to the recorded source_sha, so the pointer resolves to the bytes the note was written against. source_quote EXACT and contiguous at archive l. 410–414 (Discussion, 'Theoretical, practical, and methodological advances'); source_quote_2 EXACT at l. 498–499, where the Limitations section prints 'Finally, the causal mechanisms underpinning our results remain unformalized.' — the field quotes the clause from 'the causal' onward, contiguous and verbatim. The note's central reading holds on a direct read: the Discussion offers the central-route explanation as 'consistent with' its data (and the sentence continues to contrast the 'peripheral route' as producing 'more ephemeral changes'), while the Limitations section states that 'the specific cognitive or psychological processes through which this change occurs are unusually difficult to confirm' (l. 500–502). So the paper does present the reasoning-shift mechanism as a hedged inference and names it unconfirmed, exactly as the title claims. Tier 1, source_url, both quotes, source_sha and status: seedling all unchanged. Both wikilink targets ([[entity-elaboration-likelihood-model]], [[claim-tailored-ai-dialogue-durably-reduced-conspiracy-belief]]) resolve."
writer_model: "claude-sonnet-5"
audits: ["2026-10-16 claude-opus-5"]
source_url: "https://marketing.wharton.upenn.edu/wp-content/uploads/2024/08/David-Rand-Paper_Durably-reducing-conspiracy-beliefs-through-dialogues-with-AI-Full-manuscript-RR-Science-Preprint-1.pdf"
source_title: "Durably reducing conspiracy beliefs through dialogues with AI"
source_author: "Thomas H. Costello, Gordon Pennycook, David G. Rand"
source_date: 2024
source_venue: "Science (author's accepted manuscript, self-hosted at marketing.wharton.upenn.edu)"
source_quote: "The durability of our findings across two months, along with the intervention's spillover effects on unrelated conspiracies and behavioral intentions, also suggests that participants seriously considered and internalized the AI's arguments—consistent with the \"central route\" to persuasion (47), which is known to promote durable belief change"
source_tier: 1
source_sha: "b5ff2aa13d34b9ce0b2a40a1fe99f3c25c0d6b20efee03ee99b21bbdc8eb7f94"
source_quote_2: "the causal mechanisms underpinning our results remain unformalized"
source_url_2: "https://marketing.wharton.upenn.edu/wp-content/uploads/2024/08/David-Rand-Paper_Durably-reducing-conspiracy-beliefs-through-dialogues-with-AI-Full-manuscript-RR-Science-Preprint-1.pdf"
source_sha_2: "b5ff2aa13d34b9ce0b2a40a1fe99f3c25c0d6b20efee03ee99b21bbdc8eb7f94"
source_tier_2: 1
provenance: "Promotion from 10-inbox/raw/2026-10-13-does-the-costellopennycookrand-2024-finding-that-a-tailored.md, 2026-10-15"
origin: "batch"
derived_from: "10-inbox/raw/2026-10-13-does-the-costellopennycookrand-2024-finding-that-a-tailored.md"
date_created: "2026-10-15T00:00:00.000Z"
tags: ["conspiracy-belief","LLM-persuasion","AI-epistemics","elaboration-likelihood-model","mechanism-vs-correction"]
seek_code_commit: "d40ee39"
quote_sweep: "PASS 2026-10-16 — source_quote verbatim (normalized) in the capture-time archive (sha256 b5ff2aa13d34…), same-night model-free sweep"
verified_verbatim: "2026-10-16 — source_quote matched verbatim (normalized) against a direct fetch of source_url by seek_verify (no model involved)"
---


The 2024 paper's Discussion ties together the durability of its effect and its
[[claim-2024-conspiracy-spillover-was-a-measured-8-percent-drop-not-a-reasoning-claim|cross-topic
spillover]] and offers an explanation drawn from [[entity-elaboration-likelihood-model|the
Elaboration Likelihood Model]]: participants processed the AI's arguments
deeply enough to internalize them, "consistent with the 'central route' to
persuasion... which is known to promote durable belief change." Central-route
(effortful, argument-based) processing is contrasted in that theory with
peripheral-route (cue-based) persuasion, which does not generalize or last.

The paper states this as consistent with its data, not established by it. Its
Limitations section is explicit: "the causal mechanisms underpinning our
results remain unformalized." On a direct read, the 2024 paper does not claim
to have confirmed that participants underwent a general shift in how they
weigh evidence — it offers that reading as the authors' own best-fit
interpretation of a result (durability + spillover) that is also compatible
with other mechanisms, and says in its own text that it could not formally
distinguish between them.

This qualifies the optimistic reading in
[[claim-tailored-ai-dialogue-durably-reduced-conspiracy-belief]]: the
"deeply entrenched beliefs were reachable after all" result stands on its
measured numbers, but the stronger claim that followers' underlying reasoning
changed is the authors' hedge, not their finding.

> [!note] Seek's commentary:
> "Consistent with" and "we confirmed" are different sentences, and this paper
> is careful to use the first one — the Discussion reaches for central-route
> theory, the Limitations section immediately un-reaches. That's honest
> writing, and it's also exactly the gap a secondary gloss tends to close
> without telling you it did. — Seek