Attacks on Science Tracker

The Union of Concerned Scientists (UCS) developed an evolved methodology to track 'attacks on science'—defined as political interference in federal science—starting in January 2025. The system combines automated R scripts for RSS feed collection and keyword filtering with human content analysis to categorize attacks by type, topic, and stage of completion. The methodology aims to systematically capture a comprehensive breadth of attacks in perpetuity, including those that are only threatened, to document their chilling effects on the federal scientific enterprise.

• The Union of Concerned Scientists (UCS) defines 'attacks on science' as actions, statements, or decisions from elected officials or political appointees in federal agencies that lead to the censoring, manipulation, forging, or misinforming of scientific data, results, or conclusions produced by the government or with federal funds.
• The tracking methodology employs a two-stage automated filtering process using R scripts to identify potential attacks in news articles. The first filter screens RSS feed titles and descriptions for government/science terms, topic terms, and negative verbs; the second filter applies these same criteria to the article's full text.
• UCS categorizes attacks on science into three primary areas: attacks on federal processes, attacks on scientific research, and attacks on scientists.
• The methodology distinguishes between 'threatened' and 'completed' attacks to capture both the tangible impacts of executed attacks and the 'chilling effect' of rumored or planned interference.
• The system tracks attacks across seven specific policy topics: climate science, elections and voting, energy, environment, equity, health & safety, and a general 'any' category.
• To ensure data reliability, UCS used Gwet’s AC1 to measure intercoder reliability during training, requiring a threshold of ≥ .70 for most variables before independent coding began.
• The methodology identifies potential scientific integrity (SI) violations by automatically classifying certain attack types based on the text of the Scientific Integrity Act introduced during the 119th Congress.
• The data collection process is limited by its reliance on news coverage, which may not capture all attacks, and the potential for 'coder drift,' which UCS intends to mitigate through quarterly reliability recalculations.

APA

Barbati-Dajches, J., Ellickson, K., Velez, E., & Dietrich, A. (2026). Attacks on Science Tracker. Union of Concerned Scientists. https://www.ucs.org/sites/default/files/2026-10/Attacks-on-Science-Tracker-2-1-methodology.pdf

Chicago

Barbati-Dajches, Jules, Kristie Ellickson, Erick Velez, and Alexa Dietrich. Attacks on Science Tracker. Union of Concerned Scientists, 2026. https://www.ucs.org/sites/default/files/2026-10/Attacks-on-Science-Tracker-2-1-methodology.pdf.

Wikipedia

{{cite report |last1=Barbati-Dajches |first1=Jules |last2=Ellickson |first2=Kristie |last3=Velez |first3=Erick |last4=Dietrich |first4=Alexa |title=Attacks on Science Tracker |publisher=Union of Concerned Scientists |date=September 2026 |url=https://www.ucs.org/sites/default/files/2026-10/Attacks-on-Science-Tracker-2-1-methodology.pdf |access-date=23 October 2026 |via=Climate Insights Directory}}

BibTeX

@techreport{barbatidajches2026attacks, author = {Barbati-Dajches, Jules and Ellickson, Kristie and Velez, Erick and Dietrich, Alexa}, title = {{Attacks on Science Tracker}}, institution = {Union of Concerned Scientists}, year = {2026}, month = sep, url = {https://www.ucs.org/sites/default/files/2026-10/Attacks-on-Science-Tracker-2-1-methodology.pdf}, urldate = {2026-10-23}, note = {Indexed by Climate Insights Directory} }