Data Protection Day 2026 — Working with sdcMicro

Removing names, e-mails and birthdates is where anonymization starts, not where it ends. In this hands-on session we took a real depression dataset from 56 % re-identifiable to safely shareable — live, in the sdcMicro app — using recoding, suppression and perturbation.

Materials

Slides (PDF) Demo R script

Try it yourself in 30 seconds

install.packages("sdcMicro")
library(sdcMicro)
sdcApp()          # launches the point-and-click GUI

Everything from the demo — risk measurement, global recoding, local suppression, PRAM — is in the GUI tabs, or step by step in the R script above.

The dataset

Self-relating & depression study (n = 107, UK sample), already stripped of direct identifiers. Dataset: Colman, Vione & Kotera (2020), on PsychArchives (doi:10.23668/psycharchives.4189), from their study of psychological risk factors for depression (derailment, self-criticism, self-reassurance). Prepared copy used in the demo: coleman_prepared.csv.

Go deeper

sdcMicro on CRAN
Templ, Kowarik & Meindl (2015), Journal of Statistical Software — the sdcMicro paper (doi:10.18637/jss.v067.i04)
Meindl & Templ (2019), Algorithms — the sdcApp GUI (doi:10.3390/a12090191)
SwissAnon — anonymization services & research