A Welcome Step Toward Scientific Integrity—If Properly Implemented

The Center for Truth in Science applauds President Trump’s recent Executive Order “Restoring Gold Standard Science,” which addresses the need to improve the integrity, transparency, and reproducibility of science conducted and sponsored by the federal government. At a time when public trust in science has plummeted—with a majority of STEM researchers acknowledging a reproducibility crisis—this Executive Order addresses critical needs at the intersection of science, justice, and the economy.

The principles outlined in the order align closely with our mission: fostering science that is reproducible, transparent, collaborative, and epistemically humble and non-dogmatic. These are not merely aspirational goals—they are essential aspects of the scientific process and fundamental requirements for evidence-based regulatory and judicial systems, and for an economy that thrives on reliable innovation.

What We Support

The Executive Order gets several things right:

Transparency and data sharing: The order makes clear that agencies must comply with existing statutes and regulations requiring public disclosures of underlying data, analyses, and models (including source code). This directive promises a sea change in federal science policy that will enforce existing laws and reduce overuse of exemptions for not sharing data and models. Too often, regulatory decisions affecting billions in economic activity have relied on black-box models and unavailable data.

Acknowledging uncertainty: Science is a process, not merely a collection of facts. Scientific knowledge constantly evolves, as absolute certainty is an impossible goal. Understanding the limits of our knowledge and the potential for error are key aspects of gaining scientific knowledge. The Executive Order’s insistence on transparent documentation of data, process, and uncertainties—and how uncertainties propagate through models—represents essential requirements for honest science-based policy.

Rejecting worst-case scenarios: The order’s critique of agencies using “worst-case scenarios” like RCP 8.5 climate projections or admittedly wrong whale population models addresses a core problem. When Maine lobstermen faced industry destruction based on projections the agency itself believed were “very likely” wrong, we abandoned science for activism.

Critical Concerns That Demand Attention

However, words without enforcement are merely good intentions. Our experience suggests several areas needing clarification:

1. The “Weight of Evidence” Problem

The order’s embrace of “weight of scientific evidence” terminology raises red flags. As economist John Maynard Keynes recognized, this approach can lead to systematically biased results. Why? Because decades of publication bias mean positive findings dominate the literature while negative results languish unpublished.

For example, EPA’s Guidelines for Carcinogen Risk Assessment (EPA/630/P-03/001B, March 2005) suggest five standard hazard descriptors that do not assess the strengths and limitations of studies, including when claimed associations don’t exist. Regulatory agencies must ensure all relevant research of sufficient quality finds its way into any summary assessment of the state of the science—preventing cherry-picking.

We suggest multiple reasons for rejecting binary classification of substances as hazardous or non-hazardous. Of necessity, we are concerned about risks—quantifiable threats of cancer causation to people in real-world exposure scenarios. Cancer is a process that includes repair at low, intermittent doses. Radiation and chemicals may be carcinogens at high doses but be repaired at low doses or even beneficial.

We recommend modes of action as defined in EPA 2005 cancer guidelines (e.g., mutagenicity and target organ toxicity leading to tumor induction) as a better model for substance classification.

We also recommend systematic reviews allowing for negative and null results, not cherry-picked compilations of published positives. Established reporting guidelines exist for this (e.g., CONSORT) along with statistical tools to assess risk of bias.

2. What Is “Gold Standard” Science?

The Center believes Gold Standard Science exceeds studies conducted with integrity, transparency, and reproducibility. Gold Standard Science is not pseudoscience. Gold Standard Science follows scientific methods to stress test hypotheses, theories, and ideas, not just corroborate them. Whenever appropriate and ethically possible, Gold Standard Science focuses on dissecting and validating cause-effect relationships, not merely observations, correlations, or associations.

Sir Karl Popper instructed that actual science must be falsifiable. If a hypothesis or theory can never be proven false, it is not scientific.

What makes the scientific method work under Popper’s insights: true scientists seek to disprove their ideas, hypotheses, and theories. Only by stress testing our ideas with actual “severe” tests capable of proving them wrong when evidence points that way can we trust them sufficiently to act. If an idea withstands multiple severe tests, we gain valuable information to update our knowledge.

This explains why science constantly evolves. Using severe tests breaks us from perpetuating confirmation bias. Historically, our agencies, especially the US EPA, have been guilty of pursuing confirmation bias—identifying studies that best support an administration’s policy beliefs and preferences, regardless of quality or rigor.

3. Enforcement Mechanisms Are Unclear

Without robust enforcement mechanisms and real consequences, this Executive Order risks becoming another unfulfilled promise.

4. Model Communication Needs Teeth

While the order mentions model uncertainty, it needs stronger requirements for scenario planning. When agencies project future events, they must present multiple plausible scenarios—not just their preferred narrative. The COVID school closure guidance’s failure to use epidemiological evidence demonstrated what happens when models ignore real-world complexity and attendant uncertainty.

The Path Forward

For this Executive Order to transcend rhetoric and create real change, we recommend:

1. Statutory backing: Executive Orders can be reversed with a pen stroke. Congress must enshrine these principles in law. President Obama’s 2013 Executive Order on open data took six years to become part of statutory law.
2. Grant requirements: Federal research grants—totaling tens of billions annually—must require adherence to these standards. Bad science shouldn’t receive taxpayer funding. Grantees must acknowledge their compliance obligations.
3. Timeline for compliance: Agencies need specific deadlines to update their regulations and guidance on scientific processes and integrity, with consequences for non-compliance.
4. Protection for dissent: The order mentions protecting “dissenting viewpoints,” but whistleblowers who challenge agency science need real protection, not just promises.

Enforcement

Beyond this Executive Order, we recommend that OMB Circulars, including A-4 and Federal Acquisition Regulation, incorporate this order’s messages.

The Stakes

This isn’t merely an academic exercise. When litigation awards billions based on questionable science, destroys industries based on admittedly flawed models, or when public health decisions ignore evidence in favor of politics—real people suffer real consequences.

The Center for Truth in Science stands ready to work with the administration, Congress, and the scientific community to ensure this Executive Order becomes more than aspirational language. We’ve spent years documenting how the intersection of flawed science and the judicial system creates economic havoc. This order could be an important step toward improving how science is done and communicated by the federal government—but only with rigorous implementation and unwavering commitment to its principles.

The reproducibility crisis didn’t emerge overnight, and this Executive Order represents only a single step, albeit a very good one. Genuine and demonstrated commitment to enforcement and continuous improvement by this administration would mark a turning point in restoring both scientific integrity and public trust.

The American people deserve science that serves truth, not agendas. This Executive Order promises that standard. Now we must ensure it delivers.

The Center for Truth in Science is a nonprofit research organization dedicated to ensuring that science used in public policy and the courts is transparent, reproducible, and free from bias. We operate at the intersection of science, justice, and the economy.

Contact:
Jacob Traverse
info@truthinscience.org
truthinscience.org