Newly published research by Pan Liu examines how missing documentation can make scientific findings difficult to verify.
Scientists can save years of research data and still lose something essential: the ability to show how they reached their conclusions.
Pan Liu, Ph.D., assistant professor of communication at Marian University, examines this challenge in a newly published article on how researchers can better preserve the evidence behind their findings.
Liu's article, "What Gets Lost When Gaze Data Get Lost: Archival Reproducibility and Interpretive Accountability in Eye-Tracking Research," was published in the peer-reviewed Journal of Documentation in October 2026.
The research focuses on eye-tracking technology, which records where people look and how their eyes move. Researchers use the technology to study visual attention in fields ranging from psychology and communication to marketing and human-computer interaction.
But recording where someone looks is only the beginning. Researchers must process and interpret that information before drawing conclusions about human behavior. If those steps aren't documented and preserved, other scientists may be unable to determine how the findings were derived.
"An eye tracker can tell us where someone looked. It cannot, by itself, tell us what that person believed, wanted, or decided," Liu said. "We need to preserve not only the measurements, but also the decisions that give those measurements meaning."
Drawing on established research guidelines and an earlier eye-tracking study, Liu examines how missing or incomplete records can undermine scientists' ability to reconstruct and evaluate published findings.
His article identifies four practices that can help researchers preserve the evidence behind their conclusions:
- Keep original data intact. Preserve raw recordings separately from information that has been processed or analyzed.
- Document every step. Record how researchers turn eye movements into measurements used to support their findings.
- Distinguish observation from interpretation. Recognize that looking at something does not necessarily indicate interest, agreement, or intent.
- Preserve connections between records. Maintain documentation showing how information collected through different research methods relates to the same participants.
Together, these practices can help researchers make their findings more transparent, easier to reproduce, and open to independent evaluation.
"Reproducibility is not just about keeping files," Liu said. "It is about leaving enough of an evidence trail for someone else to understand, verify, and, when necessary, challenge how we reached a conclusion."
Although Liu's article focuses on eye-tracking research, its implications extend to other fields that rely on complex digital information. As researchers collect and analyze ever-larger amounts of data, preserving the decisions behind their conclusions is essential to maintaining scientific credibility.
The article offers a framework to strengthen research documentation and ensure scientific evidence remains useful long after a project ends.
Read the research: What Gets Lost When Gaze Data Get Lost
About the Researcher
Pan Liu, Ph.D., is an assistant professor of communication at Marian University. His research examines how people interpret, evaluate, and use communication and evidence to support judgments and decisions.