Privacy in Mobile Personalized Systems - The Effect of Disclosure Justifications
1 like544 views
The document discusses the role of disclosure justifications in mobile applications that require personal data for personalization. An experiment aimed at determining which justification types increase disclosure and user satisfaction showed that none of the justifications had the expected positive effects on disclosure behavior, perceived privacy threat, or trust in the system. The authors suggest that while justifications may not be effective, participants appreciated the help provided for disclosure decisions.
Privacy in Mobile Personalized Systems - The Effect of Disclosure Justifications
- 1. Privacy in Mobile Personalized Systems The Effect of Disclosure Justifications Bart P. Knijnenburg Alfred Kobsa Gokay Saldamli Department of Informatics, UC Irvine Department of Informatics, UC Irvine Samsung R&D Research Samsung R&D Research
- 2. Mobile apps need personal data Mobile applications often use personalization This requires personal information - Demographical data (e.g. age, hobbies, income) - Contextual data (e.g. app usage, calendar, location) INFORMATION AND COMPUTER SCIENCES
- 3. Let users control their disclosure Problem: Many people are not comfortable disclosing diverse personal information FTC, CPBoR: let users decide Privacy calculus: trade off between benefits and risks INFORMATION AND COMPUTER SCIENCES
- 4. Help users decide what to disclose Problem: This trade-off is difficult! Lack of knowledge about positive and negative consequences CPBoR: informed choice Previous research: justifications INFORMATION AND COMPUTER SCIENCES
- 5. Justification types Explain the reason why the information is requested May prove the legitimacy of the disclosure request Highlight the benefits of disclosure Privacy calculus: tip the scales in favor of the benefits Appeal to the social norm Eschew privacy calculus by conforming to the majority INFORMATION AND COMPUTER SCIENCES
- 6. Our starting point Previous work: Justifications seem to work - They increase disclosure - They increase user satisfaction -not always tested Our goal: Find out which one works best INFORMATION AND COMPUTER SCIENCES
- 7. Experiment INFORMATION AND COMPUTER SCIENCES
- 8. Experiment INFORMATION AND COMPUTER SCIENCES
- 9. Manipulations Location, etc. Gender, etc. Gender, etc. Location, etc. Context data first Demographical data first INFORMATION AND COMPUTER SCIENCES
- 10. Manipulations 5 justification types None Useful for you Number of others Useful for others Explanation INFORMATION AND COMPUTER SCIENCES
- 11. Which one is best? Which increases disclosure the most? Which increases satisfaction the most? INFORMATION AND COMPUTER SCIENCES
- 12. Results Disclosure*behavior* * Demographics*disclosure * *Context*disclosure* Context#first# Demographics#first# Context#first# Demograpics#first# 100%# 90%# 80%# 70%# 60%# 50%# 40%# 30%# 20%# 10%# 0%# INFORMATION AND COMPUTER SCIENCES
- 13. Results Disclosure*behavior* * Demographics*disclosure * *Context*disclosure* Context"first" Demographics"first" Context"first" Demograpics"first" 100%" 90%" 1" 80%" ***" 70%" *" **" *" 60%" *" *" 50%" 40%" 30%" 20%" 10%" 0%" none" useful"for"you" #"of"others" useful"for"others" explanaDon" INFORMATION AND COMPUTER SCIENCES
- 14. Results Perceived(value(of( Perceived value of Disclosure*behavior* disclosure(help( disclosure help: * Demographics*disclosure * ***" *Context*disclosure* Context"first" Demographics"first" 1,00" Context"first" ***" Demograpics"first" 100%" 3 items, e.g. “The system 0,75" **" 90%" helped 1" me to make a ***" 0,50" 80%" tradeoff between privacy 70%" 0,25" *" **" *" 60%" and usefulness” *" *" 0,00" 50%" #0,25" Higher for all except 40%" 30%" #0,50" “number of others” 20%" #0,75" 10%" #1,00" 0%" none" useful"for"you" #"of"others" useful"for"others" explanaDon" INFORMATION AND COMPUTER SCIENCES
- 15. Results Perceived(privacy( Disclosure*behavior* threat( * Perceived privacy threat: Demographics*disclosure * *Context*disclosure* Context"first" Demographics"first" 1,00" Context"first" Demograpics"first" 100%" 3 items, e.g. “The system 0,75" 90%" 1" *" 80%" has too much information ***" 0,50" 70%" about me” 0,25" *" **" *" 60%" 0,00" *" *" 50%" Higher for “useful for others” #0,25" 40%" 30%" #0,50" 20%" #0,75" 10%" #1,00" 0%" none" useful"for"you" #"of"others" useful"for"others" explanaDon" INFORMATION AND COMPUTER SCIENCES
- 16. Results Trust&in&the&& Trust in the company: Disclosure*behavior* company& * Demographics*disclosure * *Context*disclosure* 4 items, e.g. “I believe this Context"first" Demographics"first" 1,00" Context"first" Demograpics"first" 100%" company is honest when 0,75" 90%" 1" 80%" it comes ***"using the to 0,50" 70%" information I provide” 0,25" *" **" *" 60%" 0,00" *" *" Generally lower, especially 50%" $0,25" 40%" for “useful for others” 30%" $0,50" 1" 20%" $0,75" **" 10%" $1,00" 0%" none" useful"for"you" #"of"others" useful"for"others" explanaDon" INFORMATION AND COMPUTER SCIENCES
- 17. Results Sa#sfac#on)with)) Disclosure*behavior* the)system) * Demographics*disclosure * *Context*disclosure* Satisfaction with the system: Context"first" Demographics"first" 1,00" Context"first" Demograpics"first" 100%" 0,75" 90%" 6 items,1"e.g. “Overall, I’m ***" 0,50" 80%" satisfied with the system” 70%" 0,25" *" **" *" 60%" 0,00" *" *" Lower for any justification! 50%" $0,25" 40%" 30%" $0,50" 1" 20%" $0,75" **" **" 10%" $1,00" ***" 0%" none" useful"for"you" #"of"others" useful"for"others" explanaDon" INFORMATION AND COMPUTER SCIENCES
- 18. Conclusion Justifications did not have the expected effects No increase in disclosure No decrease in perceived threat, no increase in trust Satisfaction is lower ...but participants liked the disclosure help! INFORMATION AND COMPUTER SCIENCES
- 19. Reflection Why did this happen? Possible reason 1: Justifications are seen as persuasion But participants liked the disclosure help Possible reason 2: Low percentages cause disappointment Disclosure only starts to increase at around 90% for the “number of others” justification Possible reason 3: Justifications carry an implicit warning They signal that the disclosure decision is not trivial INFORMATION AND COMPUTER SCIENCES
- 20. Discussion None of our justification messages seemed to work very well Is there a “golden justification”? Different justifications may work for different types of users Has anyone tried “tailored” disclosure help? We provided objective information for privacy decisions Should we do this even if it reduces users’ satisfaction? INFORMATION AND COMPUTER SCIENCES
- 21. Thank you bart.k@uci.edu :: www.usabart.nl :: @usabart
- 22. Discussion None of our justification messages seemed to work very well Is there a “golden justification”? Different justifications may work for different types of users Has anyone tried “tailored” disclosure help? We provided objective information for privacy decisions Should we do this even if it reduces users’ satisfaction? INFORMATION AND COMPUTER SCIENCES