Log In Sign Up

Adapting Security Warnings to Counter Online Disinformation

by   Ben Kaiser, et al.

Online platforms are using warning messages to counter disinformation, but current approaches are not evidence-based and appear ineffective. We designed and empirically evaluated new disinformation warnings by drawing from the research that led to effective security warnings. In a laboratory study, we found that contextual warnings are easily ignored, but interstitial warnings are highly effective at inducing subjects to visit alternative websites. We then investigated how comprehension and risk perception moderate warning effects by comparing eight interstitial warning designs. This second study validated that interstitial warnings have a strong effect and found that while warning design impacts comprehension and risk perception, neither attribute resulted in a significant behavioral difference. Our work provides the first empirical evidence that disinformation warnings can have a strong effect on users' information-seeking behaviors, shows a path forward for effective warnings, and contributes scalable, repeatable methods for establishing evidence on the effects of disinformation warnings.


page 4

page 9

page 11


A new combination approach based on improved evidence distance

Dempster-Shafer evidence theory is a powerful tool in information fusion...

#ISIS vs #ActionCountersTerrorism: A Computational Analysis of Extremist and Counter-extremist Twitter Narratives

The rapid expansion of cyberspace has greatly facilitated the strategic ...

How Does Embodiment Affect the Human Perception of Computational Creativity? An Experimental Study Framework

Which factors influence the human assessment of creativity exhibited by ...

Giving Semantics to Program-Counter Labels via Secure Effects

Type systems designed for information-flow control commonly use a progra...

Supporting Early and Scalable Discovery of Disinformation Websites

Online disinformation is a serious and growing sociotechnical problem th...

Explanatory machine learning for sequential human teaching

The topic of comprehensibility of machine-learned theories has recently ...