Edward Snowden’s revelation about the National Security Agency’s information-gathering practices alerted the public to the nefarious uses of data collection. Stories pointing to secretive consumer tracking policies and privacy violations from companies like Facebook and Google have stirred further outcries, and education companies are not immune to criticisms either.
Yet not all data collection efforts support questionable purposes that try to peep at—or profit from—a person’s online activities. Take academia, for instance, where for decades data collected from students have been used in research that informed a broad range of education policies, from the benefits of early childhood programs to breakfast in schools, supporting student development and academic success.
These achievements explain why academic researchers are distinguishing their work from those done for commercial reasons. At a time when parents, teachers and advocacy groups are increasingly concerned about the uses of educational data, these researchers want to send a clear message about the differences between commercial and research data.
However, drawing that line may be difficult as education technology companies increasingly hire researchers to analyze data on product usage to drive further improvements—thus blurring the lines between research and commercial data collection. Proving efficacy, it seems, serves both research and commercial purposes.
“People lumped together researchers in academia [with] researchers for commercial entities that were purchasing data, and I thought, ‘Why are you doing that?” asks Dr. Kirsten Martin, an assistant professor of strategic management & public policy at the George Washington University’s School of Business, whose research focuses on online privacy. “Don’t let commercial entities fall under the name of researcher when they don’t have any of the professional obligations that we have,” she continued.
In academic research, the data collection has to be done for a project approved by university board members. He or she cannot seek to profit from the results, and cannot re-sell the data they collect. In addition, education institutions have little-to-no control over the findings, regardless of how significant (or not) they are.
That’s a far cry from the research published from companies, says Martin. It’s rare to see companies publish results that say their products are anything other than effective. Commercial research, after all, is supposed to support the goal of selling a service or product. To make these claims appear more credible, these organizations will employ academics who Martin describes as “company consultants,” noting that many of them are paid by the firms they are researching.
That relationship poses a conflict of interest that affects their ability to do objective work, she says. It also sows confusion when people see research done by professors affiliated with prominent institutions, but who don’t disclose that they are actually representing companies (not their institutions).
“You need to call them what they are. I mean, they are getting paid for their work. They don’t have any academic freedom. They can’t criticize the work, and be okay with the client,” says Martin.
“If I am an education researcher and I find that test scores are lower for certain races or ethnicities, I will not ask for permission to publish that,” Martin continues, “Academic researchers only go into a study when they know the school or company has no say over what they do.”
With commercial data collection, there are often few restrictions on how the data can be used and shared. And at times when the entity is bought and sold by other companies, the legal agreements initially put in place to restrict the uses of the data can be null and void. “There have been a couple of cases where a company can have a whole bunch of data that could be limited [in its uses], but as soon as it is bought by another company those requirements go away,” says Martin.
Other researchers point to the need to distinguish between commercial and academic data because of the risks associated with data collected by for-profit entities. At The Big Data Workshop, hosted by the National Academy of Education last August, Adam Gamoran, president of the William T. Grant Foundation, noted that commercial data, which often contains personal information, is more valuable and thus more vulnerable to ransom attacks.
“We have to understand what causes vulnerabilities, the degree of risk and the degree of incentives for hackers to be reaching out and grabbing our information,” says Gamoran. “I am skeptical that the research use of this data is creating more [incentives] for hackers... In fact, I would argue that the risk of breaches for research use of data is relatively low.”
Monica Bulger, a researcher at the Data and Society Research Institute in Manhattan, remarked at the Big Data event that education researchers have been collecting data long before the term “big data” existed. For others like her, data ethics is serious. However, she says public messaging about the importance and difference between academic and other forms of data collection has not been effective. For example, legislation that supports early childhood education programs often credits politicians who enact the policy, not the researchers who collected the data, conducted the analysis and made the recommendations to politicians.
“We don’t put up signpost when early education becomes a priority,.. saying this is brought to you by research from...” says Bulger, explaining research accomplishments that went uncredited to the audience. “We do this research in the interest of public good, and our work has led to positive change.”
Martin also warns that research is also at risk of being compromised by companies who abuse the trust put into academic researchers to further their interest. She believes districts officials should make rules requiring deals with commercial entities to be monitored by one or two outside researchers, not affiliated with the companies.
It’s a “dangerous” conflation, says Martin. “We need to really clearly say that these people are not researchers, [that] they are commercial entities doing a service for [clients],” says Martin, “It is not always clear. They should say, for example, I am here for Edutech, I am not representing Northwestern at this moment.”
