Prof. Dr. Rainer Schnell explains a method that allows for the usage of geodata in social research without raising concerns about data protection.
Information about where people work and reside is not only highly sensitive as far as data protection is concerned because it is comparatively easy to reidentify respondents of social science surveys. However, this information also comprises an enormous analytic potential: How does, for example, the spatial distance of certain institutions impact on the decisions that individuals take concerning the possibilities associated with these institutions? Hence, is there a connection between the spatial distance of educational services and their usage?
To solve this dilemma, Schnell proposes applying a distance matrix based on a randomly numbered points grid. Relying on basic laws of geometry, Schnell points out that this results in a small margin of errors when generating spatial distance measures. Simultaneously, this approach foils attempts of reidentification to the greatest degree possible as this would only be feasible with enormous effort and great mathematical or algebraic competencies. In order words: The method ensures the pseudo-anonymity of the survey respondents.
Consequently, the method–in contrast to conventionally accessible, less precise data with administrative territorial units as basic units–allows for precision and manages to balance analytic research potential on the one hand and compliance with data protection regulations on the other hand.