Instructor Notes

Overall approach

  • This module is conceptual. Each episode takes about 30 minutes plus 15 to 20 minutes for challenges.
  • It builds on the Data Management module; if learners have not taken it, spend a few minutes on file formats and on the difference between management and curation.
  • Encourage learners to curate a dataset of their own as they go: write its README in the Documentation episode, choose its repository and license in the Sharing episode, and check it against the privacy checklist in the Ethics episode.

Episode by episode

Introduction to Data Curation

  • Ask learners whether they have ever tried to reuse someone else’s data, and what was missing. The answers usually cover every episode of the module.

Metadata

  • Open the metadata of a real dataset from a state or federal portal and compare it with the list of essential elements.
  • Show the QGIS metadata editor (Layer Properties > Metadata) if QGIS is available.

Documentation

  • The README template can be handed out as a starting point for learners’ own datasets.

Preservation

  • Demonstrate a checksum: calculate one, change a single character in a text file, and calculate it again.

Sharing and Publishing Data

  • If teaching at Purdue, invite someone from the Libraries’ research data services to talk about PURR.
  • Browse ICPSR together to show how a major social science archive describes data and controls access.

Ethics and Sensitive Location Data

  • The geomasking figure uses simulated points, not real people. Emphasize that point.
  • Discussion of who could be harmed by a map works well in small groups.

Common questions

Question Short answer
“Is FAIR the same as open?” No: data can be FAIR with controlled access; the metadata stays findable
“Can I share data I downloaded?” Check its license; you can always cite it precisely instead
“Is removing names enough to anonymize data?” No, not when locations or small groups remain
“Which license should I use?” CC0 or CC BY 4.0 for most data you created; keep the original license for derived data, such as ODbL for OpenStreetMap