Verification Report: Cognitive Bias in High-Stakes Science Infographic

Date: June 25, 2026
Poster slug: bias-in-high-stakes-science
Verifier: Claude Sonnet 4.6 via verified-infographic-generator skill

Claim 1: Loss Aversion Ratio (~2:1)

Status: VERIFIED

Value used: "~2:1 — losses feel roughly twice as painful as equivalent gains"

Source:
Kahneman, D. & Tversky, A. (1979). "Prospect Theory: An Analysis of Decision under Risk."
Econometrica, 47(2), 263–291.
URL: https://ideas.repec.org/a/ecm/emetrp/v47y1979i2p263-91.html

Supporting quote:
"For some individuals, the pain from losing $1,000 could only be compensated by the pleasure
of earning $2,000" — directly illustrating the ~2:1 ratio.

Notes:
The "~" prefix is appropriate; the ratio varies by individual and domain (range ~1.5–2.5:1
across studies). The 2:1 approximation is the standard summary in the literature.
Nobel Prize in Economics 2002 awarded to Kahneman for this body of work.

Claim 2: Sunk Cost / Escalation of Commitment

Status: VERIFIED (qualitative) + DIRECTIONAL (case figures)

Value used (behavioral): "People escalate commitment to failing projects after large prior investment"

Source:
Arkes, H. R. & Blumer, C. (1985). "The Psychology of Sunk Cost."
Organizational Behavior and Human Decision Processes, 35(1), 124–140.
URL: https://www.semanticscholar.org/paper/The-Psychology-of-Sunk-Cost-Arkes-Blumer/e4564b88ca2349962a707b76be4c75076ad6bd43

Supporting quote:
"A greater tendency to continue an endeavor once an investment in money, effort, or time
has been made, despite the fact that the prior investment objectively should not influence
the decision."

Case anchor (DIRECTIONAL): "~16 years · est. $1B+"
Microsoft began Majorana research ~2010; multiple press sources report multi-billion-dollar
quantum investment. $1B+ is widely cited but not officially confirmed — "est." prefix required.

Claim 3: Confirmation Bias (Qualitative)

Status: VERIFIED (qualitative only — no percentage in poster)

Value used: "Seeking evidence that confirms what we already believe; discounting what contradicts it"

Source:
Nickerson, R. S. (1998). "Confirmation Bias: A Ubiquitous Phenomenon in Many Guises."
Review of General Psychology, 2(2), 175–220.
URL: https://journals.sagepub.com/doi/abs/10.1037/1089-2680.2.2.175
Full text: https://pages.ucsd.edu/~mckenzie/nickersonConfirmationBias.pdf

Supporting quote:
"The seeking or interpreting of evidence in ways that are partial to existing beliefs,
expectations, or a hypothesis in hand."

Notes:
No reliable single percentage for "how often" confirmation bias occurs — it is situation-
and domain-dependent. Using qualitative language only in the poster. CORRECT.

Claim 4: Publication Positive Results Rate (>80%)

Status: VERIFIED

Value used: ">80% of published papers report positive results (post-1999)"

Source:
Fanelli, D. (2010). "'Positive' Results Increase Down the Hierarchy of the Sciences."
PLOS ONE, 5(4), e10068.
URL: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0010068
PMC: https://pmc.ncbi.nlm.nih.gov/articles/PMC2850928/

Supporting quotes:
- "The proportion reporting positive results exceeded 80% after 1999, peaking at 88.6% in 2005"
- "Space Science had the lowest percentage of positive results (70.2%) and
Psychology and Psychiatry the highest (91.5%)"

Notes:
Based on 2,434 papers published 1990–2007. Use ">80%" for the infographic (the exact
88.6% peak applies to 2005 specifically; >80% is accurate for post-1999 broadly).

Claim 5: Microsoft Third-Strike Pattern

Status: VERIFIED (documented in press and Nature news)

Value used: "Three serious Nature challenges in 8 years — 2018 · 2021 · 2026"

Sources:
- 2018 retraction: Multiple press sources; Microsoft/Delft Majorana paper retracted from Nature
- 2021 retraction: Nature retraction of "Quantized Majorana conductance" paper
- 2026 critique: Legg, H. (2026). Nature "Matters Arising." DOI: 10.1038/s41586-026-10567-8
University of St Andrews press release: https://news.st-andrews.ac.uk/archive/critique-published-by-nature-challenges-microsofts-quantum-computing-claims/

Notes:
The 2018 and 2021 events were retractions; 2026 is a formal challenge (not yet retraction).
The poster will use "2018: retracted · 2021: retracted · 2026: challenged" for accuracy.

Claim 6: Motivated Reasoning (Qualitative)

Status: VERIFIED (qualitative only)

Value used: "Competitive goals shape which results scientists pursue and report"

Source:
Kunda, Z. (1990). "The Case for Motivated Reasoning."
Psychological Bulletin, 108(3), 480–498.

Notes:
Kunda established that people use reasoning to arrive at conclusions they are motivated
to reach, while maintaining an illusion of objectivity. No single percentage applicable.
Use qualitative only.

Summary Table

Claim ID	Status	Value in Poster
loss_aversion_ratio	VERIFIED	~2:1 ratio
sunk_cost_escalation	VERIFIED + DIRECTIONAL	Qualitative + "~16 yrs · est. $1B+"
confirmation_bias	VERIFIED (qualitative)	No percentage — descriptor only
publication_positive_rate	VERIFIED	>80% post-1999
microsoft_strike_count	VERIFIED	2018 · 2021 · 2026
motivated_reasoning	VERIFIED (qualitative)	No percentage — descriptor only

All 6 claims pass. 0 claims REJECTED. Pipeline proceeds to layout.

Fabrication Risk Notes

Do NOT add a percentage to confirmation bias or motivated reasoning claims.
No reliable number exists; any specific percentage would be fabricated.
The $1B+ Microsoft figure must keep the "est." prefix — it is not officially confirmed.
The 2026 event is a challenge, not yet a retraction — distinguish from 2018 and 2021.