US government coded human behavior to predict future violence

By Miriam Adler · August 10, 2026

Subjects were translated into numerical variables to evaluate their "dangerousness," but the identities of those measured remain a gap in the record.

The file's catalogue entry records a massive, disparate collection of materials. It is not the records themselves, but the scope-and-content description for a series known as the Behavioral Research Project Files, held in the US National Archives under NAID 7969903. The description reads as a checklist of a project that left a heavy paper trail: administrative files, agendas, briefing materials, correspondence, circulars, notices, drafts, meeting files, memorandums, newspapers, notes, photographs, policy and guidance files, publications, presentations, reports, studies, analyses, summaries, and transcripts.

Among these are the more specific tools of the trade: "data and coding sheets" and "case files."\ The purpose of the project, according to the archive's summary, was the pursuit of "psychological studies of the factors of criminal conduct and evaluations of dangerousness and group risk analysis."

Data and Coding Sheets

When a government project generates "coding sheets," it suggests a transition from qualitative observation to quantitative measurement. In the context of behavioral research, coding typically involves translating human behavior, speech, or reactions into numerical data for statistical analysis. The presence of these sheets alongside "transcripts" indicates that the researchers were not merely interviewing subjects, but breaking those interviews down into categorized variables.

The catalogue entry does not specify who the subjects were—whether they were prisoners, military personnel, or civilians—nor does it list the dates of the study. It only states that the project sought to understand the "factors of criminal conduct." In the mid-century era of American defense research, this kind of inquiry often overlapped with the broader goals of the Department of Defense. The public record shows that the Defense Advanced Research Projects Agency (DARPA), established in 1958, was designed to expand the frontiers of science and technology, often venturing into areas that exceeded immediate military requirements. While this specific archive entry does not explicitly name DARPA as the sponsor, the focus on "group risk analysis" fits the profile of Cold War-era strategic psychological research.

Evaluating Dangerousness

The most striking phrase in the description is the "evaluations of dangerousness." To evaluate dangerousness is to attempt to predict future violence or instability. This is a distinct task from the study of criminal conduct; the former is a predictive exercise, a claim that a person or a group can be measured and found to be a risk before an act is committed.

The records include "policy and guidance files," which suggests the research was not purely academic. Guidance files are typically used to standardize how field agents or researchers collect data and how they apply the findings of a study to real-world operations. If the goal was "group risk analysis," the research likely sought patterns that could be applied to populations—identifying the hallmarks of a "dangerous" group to better manage or neutralize that risk.

Because we are looking at a catalogue entry and not the case files themselves, the identity of the "dangerous" groups remains a gap in the record. The entry tells us the tools were used, but it does not tell us who was being measured or what the criteria for "dangerousness" were. It reports the existence of the analysis without providing the results of the analysis.

Case Files and Blue Book

In the National Archives catalogue, these behavioral files appear in close proximity to other peculiar series. A search of related records brings up the "List of Sightings" maintained by the Technical Intelligence Division and the Project Blue Book Office from 1947 to 1969, as well as sound recordings and motion picture films relating to UFO investigations.

There is no stated connection in the text between the behavioral research into criminal conduct and the investigations into unidentified flying objects. They are separate series. However, their co-location in the archive's indexing suggests a broader institutional interest in the anomalies of human experience and the limits of predictable behavior. One series tracked objects in the sky; the other tracked the volatile interior of the human mind.

Other related holdings, such as the "Behavioral Research Project Files" (NAID 7969903) and "Behavioral and Social Sciences Administrative Files" (NAID 12132119), indicate a sustained government effort to systematize the study of human behavior. The CIA Reading Room also lists collaboration on research in behavioral sciences as early as 1956.

The Behavioral Research Project Files remain a collection of boxes containing the blueprints of a psychological inquiry. The catalogue tells us that there are photographs, that there are transcripts, and that there are evaluations of risk. It does not tell us who was deemed dangerous, how they were coded, or what happened to them after the study ended.