Effect sizes in developer studies

Supervisor: Anna-Marie (ortloff@cs.uni-bonn.de)

When reporting effect sizes, most researchers classify them using arbitrary categorizations from the literature, which present fixed levels of large, medium and small effects.

Your tasks:

Assess effect sizes in a given sample of developer studies: How meaningful are the effects measured? How big is the change they lead to in real world circumstances? Plan and conduct a small (survey) study with CS / Cyber Sec students to investigate how they rate the size of effects reported in the literature


You need to have passed the lecture “Usable security and privacy” or be willing to learn the content on your own (English slides and German videos will be provided). Prior experience in planning, conducting and analyzing a human-subjects study would also satisfy the requirements.

Multiple students can work together on this topic: groups possible.