Replication: On the Ecological Validity of Online Security Developer Studies: Exploring Deception in a Password-Storage Study with Freelancers
Anastasia Danilova, Alena Naiakshina, and Johanna Deuter, University of Bonn; Matthew Smith, University of Bonn and Fraunhofer FKIE
Recruiting professional developers for studies can be challenging and one major concern for studies examining security development issues is their ecological validity—does the study adequately reflect the real world? Naiakshina et al. [CHI'19] examined the ecological validity of a password storage study conducted with students [CCS'17, SOUPS'18] by hiring freelancers from Freelancer.com. In the hope of increasing the ecologically validity, Naiakshina et al. used a deception study design wherein freelance developers were hired for a regular job using a company front created for the study, instead of openly telling the freelancers that they were taking part in a study. Based on their results, Naiakshina et al. propose the use of online freelancers to be examined further, to supplement other recruitment channels such as CS students and GitHub users. The deception in their study was used with the aim that results would reflect the real work of online freelancers. However, deception needs to be used with careful consideration, which can entail additional study design work and negotiations with ethical oversight bodies. In this paper, we take a closer look at the deception used in Naiakshina et al.’s study. Therefore, we replicate Naiakshina et al.’s work but announce and run it as a study on Freelancer.com. Our findings suggest that for this password storage study deception did not have a large effect and the open recruitment without deception was a viable recruitment method.
View the full SOUPS 2020 program at https://www.usenix.org/conference/soups2020/technical-sessions