14  Workflow 3

14.2 Decide for access restrictions

Why?

  • Some data cannot or should not be anonymized (e.g., losing their reuse potential)
    • Therefore access needs to be restricted to certain groups (as defined in consent form)
  • Some researchers fear being scooped (Laine, 2017)

restriction levels are usually limited to

  • public (everybody sees everything)
  • private (only you and your collaborators see everything)

there are different restriction levels possible for different files (in the same project). Restriction levels depend on what the research data center offers.

Level Prerequesite For what
Public Usefile anonymized data, codebooks, transcription rules
Student Usefile Short application states use purpose non-anonymized data with right to use for teaching
Scientific Usefile Longer application states use purpose, handling of data, and data analyses; identification via PostIdent non-anonymized data with right to use for research
Remote Access … + access only via virtual machine non-anonymized sensible data with right to use for research
Safe room … + access only in person at research institute non-anonymized very sensible data with right to use for research


An example: Project DESI, where

  • codebooks are publicly accessible (files on the right side)
  • video data are restricted for scientific use (files on the bottom of page)

Resources

Alternatives

  • Embargo period
    • Specify a time period, before data go public
    • Possible with research data centers and some repositories
  • Exclude certain research questions from reuse
    • Specify these research questions in the terms of use
    • Usually only possible with research data centers, except you are writing a very good license yourself
  • Create synthetic data (e.g., with R package synthpop)
    • Mimics the properties of your data
    • Then possible to share this synthetic data set

Questions to be answered at the end?
Please put them here!