Skip to content
  • From left to right students Alex Bozanic and Chase Kregor...

    Paul Aiken/ Daily Camera

    From left to right students Alex Bozanic and Chase Kregor work with Assistant Professor CaseyFiesler in her class studying the ecosystems of technology on the University of Colorado Boulder Campus on Wednesday September 13. The group was working on the Wikipedia page for Jesus and studying the listing and revision history of the page.

  • From left to right Tom Slesinger and Kyle Frye deconstruct...

    Paul Aiken/ Daily Camera

    From left to right Tom Slesinger and Kyle Frye deconstruct the posting and revision history in Assistant Professor CaseyFiesler's studying the ecosystems of technology on the University of Colorado Boulder Campus on Wednesday September 13.

  • Hannah Weber and Jacob Paul study the Wikipedia page for...

    Paul Aiken/ Daily Camera

    Hannah Weber and Jacob Paul study the Wikipedia page for Hurricane Irma in Assistant Professor CaseyFiesler's class studying the ecosystems of technology on the University of Colorado Boulder Campus on Wednesday September 13.

of

Expand
Author

What is PERVADE?

It stands for Pervasive Data Ethics for computational research. It’s a new research project funded with a $3 million National Science Foundation grant that spans six institutions, including CU Boulder. The idea is to examine how big data is gathered and used by tech companies and researchers and to develop guidelines for insuring people’s personal data isn’t mishandled even as tech companies use it to create new products.

The team includes researchers who are experts in computational science, research ethics, data practices, law and policy, health information, social computing, qualitative and quantitative research methods and data privacy.

Participants:

Dr. Katie Shilton – College of Information Studies at the University of Maryland College Park

Dr. Jessica Vitak – College of Information Studies at the University of Maryland College Park

Dr. Matthew Bietz – Department of Informatics at the University of California, Irvine

Dr. Casey Fiesler – Department of Information Science, College of Media, Communication and Information at University of Colorado Boulder

Dr. Jacob Metcalf – Data & Society Research Institute

Dr. Arvind Narayanan – Department of Computer Science at Princeton University

Dr. Michael Zimmer – School of Information Studies at the University of Wisconsin-Milwaukee

Have you tweeted, posted an Instagram photo or just generally engaged in social media before? Then it’s possible you’ve been part of a research experiment of sorts. Think that raises some ethical questions? You’re not the only one.

University of Colorado associate professor Casey Fiesler is among a group of researchers from six institutions who received a $3 million National Science Foundation grant to explore the ethics of social media research in the hopes of creating a guideline for tech companies, marketers and other researchers to follow.

The ethics questions have grown louder in the wake of several very public missteps by major tech and social media companies.

Earlier this month, a study where researchers used facial recognition technology to predict people’s sexuality was met with outrage from LGBT groups. Other headline-making incidents occurred when Facebook manipulated users’ news feeds to study users’ moods and when a researcher published 70,000 OkCupid usernames, sexual turn-ons, orientation and more.

Such incidents have made researchers, regulators and tech companies aware that some guidelines are needed.

PERVADE, or the Pervasive Data Ethics project, is designed to help iron out how technology companies, marketing gurus and academic researchers can tame the wild world of big data in ways that protect consumers and further the cause of new inventions.

The project is based at the University of Maryland and includes, in addition to CU’s Fiesler, five other researchers from Silicon Valley to New York City.

Katie Shelton is one of the lead researchers on the project who is based at the University Maryland. Shelton said the flood of data now available – from Uber trip logs, to Fitbit step counts, to Facebook likes and LinkedIn job search status markers – creates a wide-open field in which to harvest data, analyze it and develop new products and services.

“But the challenge of what personal information can be used and when is a hard one for the tech industry,” she said. “All of these devices collect information about us. Startups are always pushing the boundaries of what can be done and it raises questions. ‘If we build this new product, or if we sell our users’ data, are our users going to trust us?’ They don’t know the answer to the question.”

But academics typically have a bit more time than the average startup and for the next four years, the PERVADE team hopes to develop a sort of best practices list that tech companies and researchers can use.

A grant of this size is rare and signals a win for both ethics enthusiasts and CU’s fairly new Department of Information Science, Fiesler said.

“This is particularly exciting because it’s one of the first grants we’ve gotten within the department and it also means I’ll be able to hire and get more Ph.D. students to grow our program,” she said.

Fiesler herself is receiving a $400,000 chunk that she will put toward two focus areas: How people react to social media and whether a researcher should follow a website’s terms of service when scraping public data. To do this, she’ll be conducting interviews and surveys, as well as analyzing comments under articles and large datasets.

Researchers at universities are required to go through an ethical review process when interacting with humans, Fiesler said. But that’s not always the case if a researcher is only working with data and if people are not publicly identifiable.

“I would argue that even if you’re not working with people, if you’re working with data that was created by people, you should at least be thoughtful with how you’re using that data,” she said.

Jacob Metcalf is another PERVADE researcher. Based at University of California at Santa Cruz, Metcalf also consults with tech companies who’ve either encountered problems or who want to forestall future problems.

“Any company that is using human behavioral or social data is going to encounter serious ethics questions,” Metcalf said. With big data sets, “You can squeeze data in a way the human brain cannot…and in ways that people did not expect. That’s where we get into trouble,” he said.

“You might think your medical data is protected under HIPPA but it only covers your doctors’ notes,” he said, referring to the Health Insurance Portability and Accountability Act. “But your credit card knows what drugs you buy. What we assume is private can be inferred from non-private data sets.”

It’s common for researchers to pull from Twitter, Instagram and Reddit for their work, she said.

A lot of the times this is good, serving up a massive amount of information about human behavior for researchers to analyze, she said. For example, public health research done on Twitter can predict flu trends as well as the CDC.

“It’s just like sitting on a park bench and watching everyone around you interact,” she said. “It’s just people out there in public.”

Metcalf hopes to help bridge the distance between that public and the marketers and engineers working on new products.

“The companies I’ve encountered want to behave ethically but what they lack is the structure because we don’t train engineers to recognize that ethical decisions are analogous to complicated engineering decisions.”

Fiesler, Metcalf and other team members said their intent is not impose new rules or regulations or create impediments to innovation.

“We’re just trying to make innovation better by prompting (engineers and researchers) to ask better questions,” he said.

Business News Editor Jerd Smith contributed to this report.

Jerd Smith: 303-473-1332, smithj@dailycamera.com or twitter.com/jerd_smith