Measuring trust with science


Written by

Can math be a predictor of trust? The Office of the Director of National Intelligence’s Intelligence Advanced Research Projects Activity is trying to find out with its first challenge contest, announced Feb. 20.

The contest will sponsor the development and testing of algorithms that can use data from one participant to predict whether their partner will make trusting decisions, according to a statement from ODNI.

The contestants are then given access to data, which they test with their algorithm. The contestants must use their algorithm to submit predictions for the data.

Dubbed INSTICT, for Investigating Novel Statistical Techniques to Identify Neurophysiological Correlates of Trustworthiness, the contest will give $25,000 to the first-place contender, $15,000 to second place and $10,000 to third.

“Trust plays a fundamental role in many human relationships, organizations and behaviors,” Adam Russell, IARPA program manager, said in a statement “Knowing who can be trusted is essential for everyday interactions and is especially vital for many intelligence community missions and organizations.”

The data the contestants will be using is from IARPA’s Tools for Recognizing Useful Signals of Trustworthiness program. The program was created in 2010 and works to identify technologies and approaches to assess who can be trusted in certain conditions and contexts.

The deadline for submissions is May 5.

-In this Story-

Adam Russell, Defense Landscape, Department of Defense (DOD), Departments, Tech
TwitterFacebookLinkedInRedditGoogle Gmail