📢 Introducing RISSK: Identify at-risk interviews directly from your Survey Solutions export files

Ever wondered about the reliability of interviews in your surveys? RISSK utilizes machine learning algorithms to generate a risk score from your Survey Solutions export files. This score indicates the likelihood of unwanted interviewer behaviour in individual interviews. It is a valuable tool to prioritize suspicious interviews for verification exercises such as back-checking or audio audits.

:mag: Key Features:

  • Comprehensive. Checks many features in paradata and microdata.
  • Accessible output. Score easy to interpret, range from 0-100.
  • Generic. No adjustments or survey specifics required.
  • Flexible. Easily integratable into existing quality monitoring systems.
  • Easy to set-up and run. Only requires exports files + a few lines of command.
  • Platform independent. runs on Windows, Mac, Linux.
  • Preserves data privacy. Runs locally, no need to share/upload data.
  • Adjustable. Advanced users can tailor composition of URS.
  • Free and open source. Public repository on WorldBank GitHub.

Curious to find out more? Want to start using it? Dive in here:

1 Like