Human Biases and Privacy

This is a general topic that would study aspects related to biases that humans have regarding privacy. Many interesting things can be done here, such as studying users and their privacy preferences, focusing on a specific type of application in IoT, maybe looking at cognitive models, and focusing on biases, of course. This is a topic at the intersection of privacy and psychology.

Ask for discussions with the supervisor for more information or variations of the project. Email to arrange an appointment at the 9th floor.

Short description

Biases are manifestations of incorrect judgments resulting from cognitive tendencies that humans exhibit in situations of uncertainty or when there is insufficient information but a judgment must be made nevertheless.

The topic needs to look at how to apply methods of studying empirically the factors that affect the decisions people make in regard to privacy. A good starting source for understanding manifestations of biases and their relation to computer systems is this article.

One can either take this topic only as a literature review and theoretical analysis, or as a study involving a realistic IoT system (Internet of Things) or software engineers.

In the first case one would investigate, using empirical methods, different forms of privacy biases, maybe categorizing and exemplifying these in various ways, developing cognitive frameworks, looking at choice models and applying these to privacy.

In the second case, one can start by choosing an IoT system or an app (the supervisor can help with that) and look at the privacy that the IoT system under study provides and how the users perceive this, going further to reveal any biases involved. IoT prototypes can be created (e.g., using arduino or raspberry Pi) and used in the lab to expose biases. We would thus make a, let's call it, 'technological experiment' similar to how psychologists do, but in or case involving technological artifacts.

In both cases one would involve users and study their perception of the privacy and their behaviors with respect to privacy. We are interested in how external aspects and user characteristics contribute to forming biases in users and how these biases are being manifested.

Learning outcomes

The student will learn about privacy as well as some relevant psychology aspects. There are no relevant introductory courses at IFI as far as we know, but only some touch tangentially (e.g. IN1060, IN1050 talk about users and HCI methods, IN5010 touches on social aspects, TEK5530 , IN5600 talk about IoT systems).

However, one of the supervisors is expert in psychology, and the other in privacy; so you will be well guided with books and other reading resources.

Tags: privacy, user studies, psychology, bias, IoT, cognitive models
Published Sep. 5, 2020 12:57 PM - Last modified Sep. 5, 2020 12:57 PM

Scope (credits)