THE STORY IN BRIEF
  • Define what a row represents and how the observations were collected.
  • Give units and explain blank values before another person has to guess.
  • Test the documentation by asking a colleague to complete one small task.
MAKE TIME FOR A GOOD READ
Inside this story 2 perspectives

Imagine opening a colleague’s spreadsheet and finding a column called “duration” with the values 12, 17 and 24. Everything looks tidy. The headings line up and the numbers appear reasonable. Are those minutes or seconds? Is the clock running from the moment someone joins a queue or from the moment a member of staff starts serving them? The file cannot answer for itself.

The University of Oxford’s Bodleian iSkills programme listed an introductory open research workshop for 22 October. Its description invites researchers across disciplines to make selected parts of their work more accessible and transparent. One practical place to try that idea is a file a colleague will need to use again. A short README, a document kept with the data, can carry knowledge that might otherwise leave with the person who made the spreadsheet.

Begin with what a single row represents. In our imaginary queue study, it might represent a completed visit at a particular desk. Explain whether the count includes someone who left before being served. A colleague using the file to estimate all visitors’ waiting times needs to know which experiences made it into the observations in the first place.

Give each ambiguous value a meaning

Write down the unit for duration and the two moments used to calculate it. Then inspect the blank cells. An observed wait of zero can mean someone was helped immediately. A blank might mean the observer arrived late and missed the starting time. If a calculation treats both as zero, the result can look plausible while answering a different question.

The Open University’s data guidance recommends consistent names, descriptive folders and a way to identify versions. Putting the collection dates and a version in the supporting note helps somebody connect it to the right file. When a later version corrects a timing error, a single sentence recording the change can save the next analyst from having to compare two entire spreadsheets.

Ask somebody to use it

Give a colleague a copy they are entitled to access and set a small task. For example, ask them to find the typical recorded waiting time on one specified day and explain who their figure represents. Avoid coaching them while they work. The words they hesitate over are better evidence of missing documentation than a general question about whether the note is clear.

Finally, record any conditions on reuse and a contact for questions. Some projects can release data openly. Others involve information that can be used only within a team or under specific permissions. A readable description is valuable in either setting. Success here is modest and measurable. Another person should be able to identify the right file, interpret its columns and tell where the limits of the observations begin.

BB
ABOUT THE AUTHOR

Beyond Business Editors

Editorial desk

Beyond Business connects global developments with the decisions, ambitions and everyday lives of its readers. Our editorial desk explores the stories shaping our world.

More from this author
Keep the perspective.