One of the best Executive Orders from the Trump Administration is a recent one in regard to use of research by federal agencies. In the softer social sciences, bogus research has always been rampant. But as ideology crept into and eventually completely captured universities, the hard sciences have not been immune. Climate research is a particularly egregious example of data being manipulated and methodology slanted to get the results desired to match the ideology. Stricter rules to govern the use of research for formulating policy is long overdue. (Research EO)
The order has a number of good requirements, as it seeks to establish “Gold Standard Science”. Some are obvious, like banning falsification or fabrication of data and plagiarism. All are common sense–research should be reproducible, transparent, communicate uncertainty of results, structured for falsifiability of the hypothesis, subjected to unbiased peer review, accepting of negative results and with avoidance and/or full disclosure of any researcher conflicts of interest. All research used or considered by an agency must be fully disclosed, including any models and the source code for the models and the level of uncertainty in the study’s results.
The order could be improved. A long time ago I wrote about this issue and listed a number of requirements for research to be the basis of policy. No policy should be based on a single piece of research; there must be multiple studies on the same subject. All raw data must be revealed and every adjustment to any data used in the analyses which generated the results. Where data is adjusted in any manner, results must be shown for unadjusted data. Where a particular study design, methodology or statistical analysis technique is used; all reasonable alternatives must also be conducted and those results published. All potential sources of bias or error must be disclosed. Researchers should fully disclose any political affiliations or activity. Peer reviewers must be identified and their potential biases disclosed.
These requirements need to be extended to all government in the United States. States are even more prone to using bad research and more in need of reform. You can tell the EO is good, because “scientists” are wailing about being required to follow the standards. They know it means they can’t design the study and manipulate data to support the policy they want. Some have said they will leave the US. That would be great–take your bad research elsewhere. Publications like the Economist, once respectable but now in the full grip of TDS and wokism, decry the death of American science. More like a rebirth to real trustworthy research.
