uc berkeley / international computer science institute...41 simplifying disclosures based on...
TRANSCRIPT
![Page 1: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/1.jpg)
![Page 2: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/2.jpg)
Putting Disclosures to the Test September 15, 2016
The future of disclosures?
Moderator: Joseph Calandrino Research Director, Office of Tech. Research & Investigation, FTC
Serge Egelman UC Berkeley / International Computer Science Institute
Tamar Krishnamurti Dept. of Engineering & Public Policy
Carnegie Mellon University
Florian Schaub School of Information
University of Michigan
![Page 3: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/3.jpg)
Putting Disclosures to the Test September 15, 2016
Serge Egelman
UC Berkeley / International Computer Science Institute
![Page 4: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/4.jpg)
Serge Egelman, UC Berkeley / ICSI
improving disclosure
through contextual
integrity
Research funded by the National Science Foundation under grant CNS-1318680, and the Department of Homeland Security (contract FA8750-16-C-0140 administered by the Air Force Research Laboratory). No purchase necessary, void where prohibited, terms and conditions may apply.
![Page 5: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/5.jpg)
how often are resources
accessed in practice?
dynamic analysis
– modified Android OS and gave
phones to 36 people
– hooked all API methods invoking
permission checks
– logged contextual data surrounding
permission requests
P. Wijesekera, A. Baokar, A. Hosseini, S. Egelman, D. Wagner, and K. Beznosov. Android Permissions Remystified: A Field Study on Contextual Integrity. Proceedings of the 24th USENIX Security Symposium, 2015.
P. Wijesekera, A. Baokar, A. Hosseini, S. Egelman, D. Wagner, and K. Beznosov. Android Permissions Remystified: A Field Study on Contextual Integrity. Proceedings of the 24th USENIX Security Symposium, 2015.
![Page 6: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/6.jpg)
runtime requests?
213 requests per hour! – location (10,960/day)
– reading SMS data (611/day)
– sending SMS (8/day)
– reading browser history (19/day)
asking each time is infeasible
…but 80% wanted to block at least one request
(on average, they wanted to block 35% of all requests)
![Page 7: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/7.jpg)
what matters
expectations predicted blocking
(r=-0.39, p<0.018)
…as did app visibility
(r=0.42, p<0.001)
![Page 8: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/8.jpg)
contextual integrity
privacy is contextual
expectations drive privacy decisions
=> only prompt when access to data is likely to
be unexpected
Helen Nissenbaum, Privacy as Contextual Integrity. Washington Law Review 79, 2004 Helen Nissenbaum, Privacy as Contextual Integrity. Washington Law Review 79, 2004
![Page 9: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/9.jpg)
or better…
• automatically allow access when a user is
likely to expect it,
• automatically deny access when a user is
likely to not expect it,
• prompt when system cannot infer user
expectations (and learn from it)
![Page 10: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/10.jpg)
can we predict
privacy decisions?
field study to collect behavioral data
probabilistic prompts to measure
user expectations
![Page 11: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/11.jpg)
the results
133 Android smartphone users
176 million events recorded
4,224 prompt responses
![Page 12: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/12.jpg)
features
permission information – permission
– visibility
– time of day
user behavior – browsing habits
– audio preferences
– screen locking habits
past decisions – under different visibility levels
– under different foreground applications
![Page 13: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/13.jpg)
challenging the status quo
Error Rate Average Prompts/User
Ask-on-first-use 19.47% 12.34
ML Model (behavior-only)
18.82% 0.00*
ML Model 4.27% 25.60
ML Model (low-prompt)
12.67% 12.46
![Page 14: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/14.jpg)
open questions
what is an acceptable accuracy level?
what are the legal issues?
how can this be applied in other domains?
![Page 15: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/15.jpg)
Putting Disclosures to the Test September 15, 2016
Tamar Krishnamurti
Dept. of Engineering & Public Policy Carnegie Mellon University
![Page 16: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/16.jpg)
A Patient-Centered Approach to
Informed Consent
Tamar Krishnamurti, PhD Assistant Research Professor
Carnegie Mellon University
The research presented was funded by ICON plc and the Swedish Foundation for
Humanities and Social Sciences (Riksbankens Jubileumsfond) Program on Science
and Proven Experience.
![Page 17: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/17.jpg)
What is “informed” consent?
17
Krishnamurti, T. & Argo, N. A Patient-Centered Approach to Informed Consent: Results from a Survey and Randomized Trial. Medical Decision Making 2016 Aug;36(6):726-40.
![Page 18: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/18.jpg)
Existing standards
• ICH Good Clinical Practice (GCP)
• Dept. of Health and Human Services
federal regulations, e.g.
– The purpose of the trial
– The trial treatment(s)
– Random assignment
– The reasonably expected benefits
– Participation is voluntary etc. etc.
18
![Page 19: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/19.jpg)
New technologies, new options
19
![Page 20: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/20.jpg)
Research Questions
1. What consent information do patients care about ?
2. Can we generate a patient-centered consent form that
meets normative guidelines?
3. Can these be delivered in different media?
4. Are patient-centered consent forms at least as good
as traditional consent forms?
20
![Page 21: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/21.jpg)
Mturk sample
• 118 Asthma patients
• Age range 21-61years ; 44% female
• Randomly assigned to 4.5 page excerpt from 17 page
clinical trial informed consent document
• Embedded attention checks:
– 82% successfully completed at least one of the two
21
![Page 22: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/22.jpg)
Consent Priority Selection
22
![Page 23: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/23.jpg)
23
Consent Priority Rating
![Page 24: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/24.jpg)
24
% of people selecting a
specific sentence
% of people
selecting specific
concepts
Conceptual category of
specific sentence
Automatic consent form generator
![Page 25: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/25.jpg)
Resulting patient-designed form
25
![Page 26: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/26.jpg)
Did it meet normative criteria?
– The trial treatment(s) and random assignment
– The trial procedures to be followed
– The reasonably foreseeable
– The reasonably expected benefits
– Participation is voluntary etc., etc.,
26
![Page 27: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/27.jpg)
27
New technologies, new options
![Page 28: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/28.jpg)
Lab-based Evaluation
28
• 76 Asthma patients
• Age range 20-63 ; 54.3% female
• Randomly assigned to patient-centered (written or video)
or traditional consents
![Page 29: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/29.jpg)
29
Kno
wle
dge S
core
123456789
101112
P-C paper P-C video Traditional
Note: bars show standard errors
P-C V vs. T
P =.80, d =.58 P-C V vs. T
P =.80, d =.58 P-C V vs. T
P =.80, d =.58 P-C V vs. T
P =.80, d =.58
No lost knowledge with patient-centered(P-C) formats
![Page 30: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/30.jpg)
Strongly
disagree
Strongly
agree
How much do you agree with the following statement:
The benefits of this clinical trial outweigh the risks 30
1
2
3
4
5
P-C paper P-C video Traditional
No difference in perceived risks or benefits
![Page 31: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/31.jpg)
Not at all
engaged
Completely
engaged
P =.01, d = .72
P =.06, d = .57
How engaged were you in reading the consent form? 31
1
2
3
4
5
P-C paper P-C video Traditional
Patient-derived formats are more engaging
![Page 32: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/32.jpg)
Take-aways and next steps
32
• Greater engagement in patient-centered consent with
large effect sizes
• No differences found in critical decision factors
• Open questions include:
• Mturk reliability
• How does affect, type of risk, chronicity of disease etc.
play a role
![Page 33: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/33.jpg)
Thank you
Contact information
Tamar Krishnamurti
Carnegie Mellon University, 129 BH
Pittsburgh, PA 15213
Website: https://www.cmu.edu/epp/people/faculty/tamar-krishnamurti.html
33
![Page 34: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/34.jpg)
Putting Disclosures to the Test September 15, 2016
Florian Schaub
School of Information University of Michigan
![Page 35: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/35.jpg)
Contextualizing and Personalizing
Privacy Notices and Controls
Florian Schaub
Putting Disclosures to the Test
Federal Trade Commission
Sept. 15, 2016
The research presented was funded in part by the National Science Foundation, the Defense Advanced Research Projects Agency,
the Air Force Research Laboratory, Google, Inc., Yahoo! Inc., and the Carlsberg Foundation.
![Page 36: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/36.jpg)
36
privacy policies are too complex
![Page 37: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/37.jpg)
37
privacy policies are too complex
how can we reduce information overload and
enable informed privacy decision making?
how can we reduce information overload and
enable informed privacy decision making?
![Page 38: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/38.jpg)
38
simplifying disclosures based on expectations
38
![Page 39: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/39.jpg)
39
simplifying disclosures based on expectations
39
J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal. How Short is Too Short? Implications of Length and Framing on the
Effectiveness of Privacy Notices. Symposium on Usable Privacy and Security 2016.
privacy policy
www.fitbit.com/
legal/privacy-policy
3,500 words
layered notice
www.fitbit.com/privacy
1,300 words
![Page 40: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/40.jpg)
40
simplifying disclosures based on expectations
40
J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal. How Short is Too Short? Implications of Length and Framing on the
Effectiveness of Privacy Notices. Symposium on Usable Privacy and Security 2016.
our compact disclosure format
![Page 41: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/41.jpg)
41
simplifying disclosures based on expectations
41
J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal. How Short is Too Short? Implications of Length and Framing on the
Effectiveness of Privacy Notices. Symposium on Usable Privacy and Security 2016.
determine privacy expectations / awareness of data practices
• online survey with amazon mechanical turk (n=70)
• participants asked to look at a specific fitness wearable
• rate likelihood of certain data collection and
sharing practices
• actual practices mixed in with fictitious practices
![Page 42: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/42.jpg)
42
simplifying disclosures based on expectations
42
J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal. How Short is Too Short? Implications of Length and Framing on the
Effectiveness of Privacy Notices. Symposium on Usable Privacy and Security 2016.
determine privacy expectations / awareness of data practices
true true true true false
false false false false
![Page 43: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/43.jpg)
43
simplifying disclosures based on expectations
43
J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal. How Short is Too Short? Implications of Length and Framing on the
Effectiveness of Privacy Notices. Symposium on Usable Privacy and Security 2016.
baseline expectations / awareness
0 10 20 30 40 50 60 70 80 90 100
StepsDistance
Info Posted to ProfileWhen Exercising
HeartrateStairs Climbed
NameSleep
Exercise Comp. to FriendWeightHeight
Location SpecificFitbit Friends
Companies Providing ServicesDirected Organizations
GovernmentWhere to Find Privacy Policy
Use Fitbit Without an AccountSelling Data Conditions
Data Retention Policy
Sharing with Sharing w
Sharing with Sharing with Collection of Location (Specific)
Collection of Collection of
Collection of Collection of
Collection of
Collection of Collection of
Collection of Collection of
Collection of
Mis
c.
Sh
are
C
ollecti
on
Sharing with
Collection of
% correct
![Page 44: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/44.jpg)
44
simplifying disclosures based on expectations
44
J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal. How Short is Too Short? Implications of Length and Framing on the
Effectiveness of Privacy Notices. Symposium on Usable Privacy and Security 2016.
baseline expectations / awareness
0 10 20 30 40 50 60 70 80 90 100
StepsDistance
Info Posted to ProfileWhen Exercising
HeartrateStairs Climbed
NameSleep
Exercise Comp. to FriendWeightHeight
Location SpecificFitbit Friends
Companies Providing ServicesDirected Organizations
GovernmentWhere to Find Privacy Policy
Use Fitbit Without an AccountSelling Data Conditions
Data Retention Policy
Sharing with Sharing w
Sharing with Sharing with Collection of Location (Specific)
Collection of Collection of
Collection of Collection of
Collection of
Collection of Collection of
Collection of Collection of
Collection of
Mis
c.
Sh
are
C
ollecti
on
Sharing with
Collection of
85%
![Page 45: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/45.jpg)
45
simplifying disclosures based on expectations
45
J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal. How Short is Too Short? Implications of Length and Framing on the
Effectiveness of Privacy Notices. Symposium on Usable Privacy and Security 2016.
exclude most expected practices (85%)
full medium (85%)
![Page 46: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/46.jpg)
46
simplifying disclosures based on expectations
46
J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal. How Short is Too Short? Implications of Length and Framing on the
Effectiveness of Privacy Notices. Symposium on Usable Privacy and Security 2016.
baseline expectations / awareness
0 10 20 30 40 50 60 70 80 90 100
StepsDistance
Info Posted to ProfileWhen Exercising
HeartrateStairs Climbed
NameSleep
Exercise Comp. to FriendWeightHeight
Location SpecificFitbit Friends
Companies Providing ServicesDirected Organizations
GovernmentWhere to Find Privacy Policy
Use Fitbit Without an AccountSelling Data Conditions
Data Retention Policy
Sharing with Sharing w
Sharing with Sharing with Collection of Location (Specific)
Collection of Collection of
Collection of Collection of
Collection of
Collection of Collection of
Collection of Collection of
Collection of
Mis
c.
Sh
are
C
ollecti
on
70%
Sharing with
Collection of
![Page 47: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/47.jpg)
47
simplifying disclosures based on expectations
47
J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal. How Short is Too Short? Implications of Length and Framing on the
Effectiveness of Privacy Notices. Symposium on Usable Privacy and Security 2016.
compact disclosures
full medium (85%)
short (70%)
![Page 48: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/48.jpg)
48
simplifying disclosures based on expectations
48
J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal. How Short is Too Short? Implications of Length and Framing on the
Effectiveness of Privacy Notices. Symposium on Usable Privacy and Security 2016.
compact disclosures
full medium (85%)
short (70%)
![Page 49: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/49.jpg)
49
simplifying disclosures based on expectations
49
J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal. How Short is Too Short? Implications of Length and Framing on the
Effectiveness of Privacy Notices. Symposium on Usable Privacy and Security 2016.
testing the compact disclosures
• online survey with amazon mechanical turk (n=400)
• similar design as baseline survey
• but after looking at specific fitness wearable, participants see one
of the compact disclosures
• plus control condition without disclosure (same as baseline)
![Page 50: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/50.jpg)
50
simplifying disclosures based on expectations
50
J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal. How Short is Too Short? Implications of Length and Framing on the
Effectiveness of Privacy Notices. Symposium on Usable Privacy and Security 2016.
testing the compact disclosures
findings
• participants who saw disclosure had significantly higher awareness
of practices (% correct)
![Page 51: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/51.jpg)
51
simplifying disclosures based on expectations
51
J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal. How Short is Too Short? Implications of Length and Framing on the
Effectiveness of Privacy Notices. Symposium on Usable Privacy and Security 2016.
testing the compact disclosures
findings
• participants who saw disclosure had significantly higher awareness
of practices (% correct)
• similar awareness with medium and full disclosures (no sign.
diff.), but significant drop in awareness with short disclosure
![Page 52: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/52.jpg)
52
simplifying disclosures based on expectations
52
J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal. How Short is Too Short? Implications of Length and Framing on the
Effectiveness of Privacy Notices. Symposium on Usable Privacy and Security 2016.
testing the compact disclosures
findings
• participants who saw disclosure had significantly higher awareness
of practices (% correct)
• similar awareness with medium and full disclosures (no sign.
diff.), but significant drop in awareness with short disclosure
• no difference in time spent on disclosure – regardless of length
![Page 53: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/53.jpg)
1. emphasize likely unexpected or surprising information
53
simplifying privacy notices and controls
53
F. Schaub, R. Balebako, A.L. Durity, L.F. Cranor. A Design Space for Effective Privacy Notices. Symposium on Usable Privacy and Security 2015.
F. Schaub. B. Könings, M. Weber. Context-adaptive Privacy: Leveraging Context Awareness to Support Privacy Decision Making, IEEE Pervasive
Computing, vol. 14(1), 2015.
![Page 54: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/54.jpg)
1. emphasize likely unexpected or surprising information
2. contextualize information based on type of service, user
activity and user goals
54
54
F. Schaub, R. Balebako, A.L. Durity, L.F. Cranor. A Design Space for Effective Privacy Notices. Symposium on Usable Privacy and Security 2015.
F. Schaub. B. Könings, M. Weber. Context-adaptive Privacy: Leveraging Context Awareness to Support Privacy Decision Making, IEEE Pervasive
Computing, vol. 14(1), 2015.
simplifying privacy notices and controls
![Page 55: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/55.jpg)
1. emphasize likely unexpected or surprising information
2. contextualize information based on type of service, user
activity and user goals
3. personalize information based on user characteristics and
individual information needs
55
55
F. Schaub, R. Balebako, A.L. Durity, L.F. Cranor. A Design Space for Effective Privacy Notices. Symposium on Usable Privacy and Security 2015.
F. Schaub. B. Könings, M. Weber. Context-adaptive Privacy: Leveraging Context Awareness to Support Privacy Decision Making, IEEE Pervasive
Computing, vol. 14(1), 2015.
simplifying privacy notices and controls
![Page 56: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/56.jpg)
56
personalized privacy assistants
56
learning privacy profiles from users’ privacy settings
B. Liu, M. Andersen, F. Schaub, H. Almuhimedi, S. Zhang, N. Sadeh, A. Acquisti, Y. Agarwal. Follow My Recommendations: A Personalized Privacy
Assistant for Mobile App Permissions. Symposium on Usable Privacy and Security 2016.
84 Android users (rooted phones)
2 week field study (1 nudge per day)
![Page 57: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/57.jpg)
57
personalized privacy assistants
57
providing personalized privacy settings recommendations
B. Liu, M. Andersen, F. Schaub, H. Almuhimedi, S. Zhang, N. Sadeh, A. Acquisti, Y. Agarwal. Follow My Recommendations: A Personalized Privacy
Assistant for Mobile App Permissions. Symposium on Usable Privacy and Security 2016.
profile
assignment
permission
recommendations
![Page 58: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/58.jpg)
58
personalized privacy assistants for internet of things
58
www.privacyassistant.org
• aggregate disclosures and
controls across IoT systems
• context-aware privacy
decision support and
configuration
• personalized
recommendations and
adaptation
![Page 59: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/59.jpg)
59
personalized privacy assistants for internet of things
59
www.privacyassistant.org
• aggregate disclosures and
controls across IoT systems
• context-aware privacy
decision support and
configuration
• personalized
recommendations and
adaptation
• machine-readable privacy
disclosures and controls
needed
![Page 60: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/60.jpg)
60
summary
60
• emphasize unexpected / surprising practices in disclosures
• adapt disclosures to specific contexts
• personalize disclosures and controls
• need for machine-readable disclosures and controls
![Page 61: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/61.jpg)
61
summary
61
• emphasize unexpected / surprising practices in disclosures
• adapt disclosures to specific contexts
• personalize disclosures and controls
• need for machine-readable disclosures and controls
• online studies effective for eliciting expectations and testing
disclosure variants
• additionally lab and field studies under real conditions
![Page 62: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/62.jpg)
62
summary
62
• emphasize unexpected / surprising practices in disclosures
• adapt disclosures to specific contexts
• personalize disclosures and controls
• need for machine-readable disclosures and controls
• online studies effective for eliciting expectations and testing
disclosure variants
• additionally lab and field studies under real conditions
Florian Schaub [email protected]
The research presented was funded in part by the National Science Foundation, the Defense Advanced Research Projects Agency,
the Air Force Research Laboratory, Google, Inc., Yahoo! Inc., and the Carlsberg Foundation.
![Page 63: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/63.jpg)
Putting Disclosures to the Test September 15, 2016
The future of disclosures?
Moderator: Joseph Calandrino Research Director, Office of Tech. Research & Investigation, FTC
Serge Egelman UC Berkeley / International Computer Science Institute
Tamar Krishnamurti Dept. of Engineering & Public Policy
Carnegie Mellon University
Florian Schaub School of Information
University of Michigan
![Page 64: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/64.jpg)
![Page 65: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/65.jpg)
Putting Disclosures to the Test September 15, 2016
Jessica Rich
Director, Bureau of Consumer Protection, FTC
![Page 66: UC Berkeley / International Computer Science Institute...41 simplifying disclosures based on expectations 41 J. Gluck, F. Schaub, A. Friedman, H. Habib, N. Sadeh, L.F. Cranor, Y. Agarwal](https://reader034.vdocument.in/reader034/viewer/2022051813/6033d39ef4278060a31c6823/html5/thumbnails/66.jpg)