Stanford University scientists are embarking on a multidisciplinary effort to cut through one of the toughest and most persistent obstacles in academic research: red tape.
The issue concerns the growing number of instances in which researchers seek access to large datasets – often from private companies, but also from the government – to conduct analyses in fields such as human health, behaviour, economics and more.
The routine negotiations that are often necessary to obtain the data do involve legitimate concerns such as human subject protections, legal liability, secure data storage procedures and corporate rights to results, a team of Stanford scientists acknowledged.
However, the team’s interviews with officials at nearly all the 50 largest research universities in the US showed that many of the most time-consuming elements of the process involved repeatedly rehashing relatively standard elements of those requirements, which could be made far more uniform and routine.
“Why do the answers to those questions have to be different every time?” said one of the project leaders, Michelle Mello, a professor of law and medicine at Stanford.
“I will virtually guarantee you that if you talk to any working social scientists in the US today who rely on secondary data and you ask about this topic, they will groan wearily,” Professor Mello continued. “We have all been through it.”
A paper published in Science by Professor Mello and colleagues says that one option is the creation of a universal template for data use agreements. But the eventual solution to data bottlenecks, Professor Mello said, would require negotiation over some key understandings, most fundamentally the insistence of many companies on a right to withhold research findings – a demand that is generally unacceptable in academia.
Areas where common templates might be easier to reach, Professor Mello continued, involved cases in which companies with very low-risk data reflexively made costly demands for extensive data protection systems. Some researchers, she said, have had to cancel projects simply because their universities could not provide such unnecessarily elevated levels of security.
Efforts to standardise data-sharing agreements seem worth making, although researchers should understand that the information is often valuable and that companies currently have little incentive to share it, said Andrey Fradkin, an assistant professor of business at Boston University.
“For most companies, data sharing with academic researchers is not a high priority,” Dr Fradkin said. “Companies with specific problems typically prefer to consult with researchers with an exchange of money in a way that does not concern publishable research.”
The Science paper says that potential solutions for universities include increasing staffing in the offices responsible for negotiating data use agreements, and improving coordination between such offices. Other options include educating academics about the complexity and importance of such agreements so they can be clearer at the start of the negotiation process about the nature of their research and the data on which it focuses.
“Long waits for data are already a fact of life,” Professor Mello said. “And the problem will only grow with the inexorable increase in the volume and complexity of data exchanges.”
