Beyond Data Forensics- How to Deal with Flagged Candidates

The use of data forensics in high-stakes testing, such as licensure testing, is a well-established practice. Empirical data, processed in an unbiased manner, is an effective way of detecting or confirming examination misconduct. Unfortunately, the algorithms used to flag candidates are often complex and poorly understood by decisions makers. Consequently, regulators often struggle with how to proceed once candidates are flagged, especially in cases where there is no external corroborating evidence. A healthcare regulatory body encountered a situation where several candidates were flagged through routine data forensics. This was the first time this regulator had to deal with candidates flagged during data forensics with no pre-existing corroborating information.

The evaluation of this flagged result required careful legal and regulatory expertise to ensure 1) the integrity of the entry-to-practice vetting process and 2) due process for the candidate all while working within the parameters of the regulatory authority. The purpose of this presentation is to provide introductory information on data forensics, share insight and lessons learned, and provide guidance to other regulators on how to proactively plan for this scenario should they encounter it in the future.

Speakers:
Chris Beauchamp, Vice President, Psychometrics, Meazure Learning

Melanie Therrien, Deputy Registrar, College of Licensed Practical Nurses of Alberta