A Privacy-Preserving Method for Longitudinal Participant Linkage in Web Surveys
A Privacy-Preserving Method for Longitudinal Participant Linkage in Web Surveys

Summary/Abstract: Aim: To enable longitudinal linkage in online panel surveys without collecting direct identifiers andwhile aligning with modern data-protection requirementsDesign / Research methods: The article proposes a client-side protocol where participants create areproducible secret from a self-chosen pseudonym and an ordered image sequence. The browsernormalizes and cryptographically hashes these inputs to derive a short alphanumeric core code, adds amodulus-97 checksum for strict local validation, and the backend stores only a salted hash scoped to aspecific study (form-family) context.Conclusions / findings: This paper introduces a client-side protocol for generating anonymous yetlinkable participant identifiers in web-based surveys by deriving a reproducible code from a user-chosenpseudonym and image sequence entirely in the browser, and by storing only a form-family–salted hashon the server for longitudinal linkage within a study. The design incorporates a checksum for strict client-side validation and is intended to reduce spurious identifiers caused by typographical errors; empiricalvalidation of matching performance, usability, and security properties is left for future work.Originality / value of the article: The work refines SGIC-style respondent-generated linkage bycombining graphical secrets with browser-based cryptographic processing, checksum-based client-sidevalidation, and form-family salting-yielding a concrete, implementable algorithm that improves privacy-respecting longitudinal linkage.

• Details
• Contents