Data Horror Stories: What Haunts Projects and How to Exorcise It

This month, we hosted a three-part series co-sponsored by Countway Library and the HMS/HSDM Office for Postdoctoral Fellows.

đŸ˜± Some research nightmares begin with a click, a crash, or a file named final_FINAL_reallyfinal.

Every researcher has a horror story: the hard drive that fails the night before a deadline; the dataset left behind by a departing lab member, cryptic and undocumented, like a message written in a language no one remembers; the folder containing seventeen versions of a file, each one labeled “final.”

These aren’t just cautionary tales. They’re real, they’re common, and many of the risks can be prevented—or at least made less frightening—with thoughtful data-management practices.

In our first two webinars, we shared stories of data management gone wrong, explored why these problems happen, and considered how they affect research teams. The series concluded with practical tools and resources to help researchers move from ghoul to G.O.A.T.—turning data-management fears into sustainable practices for more secure, usable, and reproducible research.

🧰 The stories may be spooky, but the solutions don’t have to be: start small, ask for help, and keep your data from becoming a ghost of projects past.

Did you miss out? You can explore resources from three events!

Event 1: When Your Data Ghosts You

đŸ‘» Good documentation keeps important project knowledge from vanishing into the fog.

The first event explored what happens when research materials disappear, become impossible to find, or lose their meaning. Stories of failed backups, confusing file structures, competing “final” versions, and undocumented datasets showed how easily valuable work can become inaccessible—even when the files still exist.

Takeaway: Protect your work with reliable backups, clear file and folder organization, sensible version control, and documentation that helps someone else understand what they’re looking at. A README may not be glamorous, but it can keep your data from becoming a ghost story.

Presentation Slides for When Your Data Ghosts You

Webinar Recording of When Your Data Ghosts You

Event 2: The Darkside of Data

đŸȘŠ Access reviews help make sure former team members don’t retain a key to the crypt.

The second event turned to security, access, compliance, and sharing. The stories showed how unsecured devices, lingering access, publication restrictions, and informal data sharing can put sensitive information—or research plans—at risk.

Takeaway: Know where your data are, who can access them, and what rules apply before sharing or publishing. When you’re uncertain, pause and ask the appropriate institutional resource. It’s much less spooky to ask a question early than to discover a problem after the fact.

Presentation Slides for The Darkside of Data

Event 3: From Ghoul to G.O.A.T.

🧟 No crystal ball required: ask questions early, use available resources, and make your data easier to find, understand, and reuse.

In the final event, we moved from the “what went wrong?” to the “what can we do next?” We had lively discussions over lunch, and panelists shared practical advice on storage and backups, organization, documentation, security, compliance, and available institutional support.

Takeaway: You don’t need to transform your entire workflow overnight. Start with a small, feasible step—a README, a storage-location index, a backup check, or a clearer folder structure—and build from there. As your data and projects change, your practices can evolve too. And you don’t have to figure it all out alone: ask questions and make use of resources such as OSRDS, ORARC, HLC Research Data Compliance, FASRC, and project onboarding checklists.

Presentation Slides for From Ghoul to G.O.A.T.

Resources to Avoid a Data Horror Story

Harvard University Resources

Additional Data Horror Stories