When I think back to our 2024 NSF-funded workshop on research security, one thing stands out. We had a lot of important conversations, but we were missing something essential: data.
Everyone was talking about research security. We discussed foreign influence, intellectual property, undisclosed relationships and the ways adversarial nations might seek to exploit America’s open research environment. But when we asked a simple question — “How often is this actually happening?” — we found ourselves facing a massive data vacuum.
That is a problem.
Our university’s ability to lead technological innovation depends on an open and collaborative research environment. International partnerships, the exchange of ideas and the free flow of knowledge are fundamental to the way modern research works. At the same time, that openness can create opportunities for exploitation, coercion, theft and other security risks.
Federal policies, including NSPM-33 and the CHIPS and Science Act, have made research security a national priority. But priorities need evidence behind them. Without reliable data, universities and policymakers are left to make decisions based largely on anecdotes. That can lead us in two very different directions: strategies that overstate the threat and create unnecessary burdens, or strategies that underestimate real vulnerabilities.
My colleagues and I hope to change that, and with our recent award from the National Science Foundation, Rice will be the first university to study the nature, scale and scope of foreign influence in institutions of higher education.
Protecting Privacy to Uncover the Truth
The first step in addressing a vulnerability is being able to measure it. But research security presents a unique challenge: People may be reluctant to answer sensitive questions honestly if they fear professional consequences.
If we want researchers to tell us what is really happening, we have to create an environment where they can do so safely.
To help us achieve that, we will use an indirect questioning approach known as the Crosswise Model. The Crosswise Model mathematically decouples individual responses from the aggregate results. That means we can learn about patterns across a population without being able to connect a particular answer back to a particular person.
We are also taking a rigorous approach to ensuring our findings represent the diversity of the academic research enterprise. Our national sample will be stratified according to university research expenditures and disciplinary focus, including STEM and non-STEM institutions and fields.
Turning Evidence into Action
What we learn could have implications well beyond this study.
For Rice, this is an opportunity to help build a stronger approach to research security — one that protects research and intellectual property while preserving the openness that makes academic discovery possible.
Research security should not mean closing ourselves off from responsible international collaboration. It should mean understanding where the real risks are and responding to them thoughtfully.
Better data can help us do both.