The example that I have chosen to discuss here is not a conventional experiment but is definitely one related to information. Datta, Tsachantz and Datta (2015) who carried out the project and wrote the paper called it “Automated Experiments on Ad Privacy Settings: A Tale of Opacity, Choice, and Discrimination”. Google had come out with Ad settings as a way to allow users exercise some control over the profiles Google creates on users, and Datta, Tschantz and Datta (2015) wanted to find out how user behaviors, Google's ads and Ad Settings interact, and created an automated tool called AdFisher that can run browser based experiments and can conduct statistical analysis.
The main idea that is that when a user changes settings in Ad settings, they also exert some choice over the profiles Google creates on users and the search results should reflect this change.
Datta, Tsachantz and Datta (2015) determined that there were three privacy properties that could support their methodology for finding causal relationships and they were: 1) Opacity, 2) Choice, and 3) Discrimination. They created two groups for statistical analysis, one of which receives experimental treatment while the other remains a control, and other variables were changed accordingly to the property they were analyzing. I will describe their experimental variables for analyzing opacity below:
Null hypothesis: The two groups with the same ad settings receives ads from the same distribution despite being subjected to different experimental treatments
Independent variable: Visiting websites associated with certain interests
Dependent variable(s): Interests listed on Ad settings pages and the ads being displayed
Controlled variable: The control group did not visit the related sites
No comments:
Post a Comment