Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Watch this 12-minute conference presentation examining critical flaws in statistical methodology within usable privacy and security research. Learn about a comprehensive systematic review of 121 papers from the Symposium on Usable Privacy and Security spanning 20 years, revealing widespread issues with null hypothesis significance testing (NHST) implementation. Discover how 23% of analyzed papers inadequately account for non-independence between samples, potentially invalidating their claims, while 58% lack sufficient information to verify their statistical assertions. Explore common problems including imprecise specification of statistical tests, minor statistical validity issues, and reporting practices that deviate from established standards. Gain insights into the researchers' methodology for coding statistical tests and identifying validity, reporting, and interpretation problems that undermine scientific assertions. Understand the implications of these findings for the field and receive practical recommendations for improving statistical reporting and statistical thinking in usable privacy and security research.
