DARPA Wants a ‘Social Supercollider’ To Help It Understand Humans

The Pentagon wants to fundamentally change how social science research is conducted.

The Defense Department uses principles of human behavior to make major decisions affecting national security, but social science is inherently limited, according to the Defense Advanced Research Projects Agency, mostly because social scientists "rarely, if ever, have ground truth" about "actual causes of observed behaviors in the 'real world.'"

To help mitigate those shortcomings, DARPA is gathering information on ways, and potentially new technology, that can assess how closely social science research methods do represent the real world. Simulations, for instance, could be used to calibrate whether the inferences about human behavior social scientists make when they're conducting an analysis are valid.

» Get the best federal technology news and ideas delivered right to your inbox. Sign up here.

More specifically, DARPA is looking for "social supercollider”: a system that could be a sort of test-bed for research methods and tools, possibly preventing decision-makers from drawing conclusions about human behavior from fundamentally flawed studies.

Without such a system, decision-makers can't fully understand "different research methods’ limitations in being able to correctly identify and characterize the causes" of behaviors, the request for information said.

A successful social supercollider might help scientists determine whether their methods are powerful enough to understand complex human behavior, the RFI said. With such a system, they may also be able to test out whether their inferences about human behavior are accurate, "with precision and certainty almost never available in the 'real world.'"

An effective test-bed might also let scientists come up with totally new "hybrid social science research methods" by testing their precision virtually, the RFI said.