Transcript
Digital Rights and Freedoms: A Framework for Surveying Users and
Analyzing Policies
Todd Davies SocInfo 2014
Privacy Data Portability CreaBve Commons Free SoCware Open Source ParBcipatory Design Wikipedia Democracy
Right to Connect Accessibility Freedom of Expression Open Access Right to Be ForgoJen Net Neutrality Open Protocols
WebWeWant.org from the Mission:
“create a world where everyone, everywhere is online and able to parBcipate in a free flow of knowledge, ideas, collaboraBon and creaBvity over the open Web.”
Webwewant.org 5 Key Principles
1. Freedom of expression online and offline 2. Affordable access to a universally available
communicaBons plaUorm 3. ProtecBon of personal user informaBon and
the right to communicate in private 4. Diverse, decentralised and open
infrastructure 5. Neutral networks that don’t discriminate
against content or users
1. Privacy Control
The user is able to know and to control who else can access their data.
Associated concepts: • Originator-‐discreBonary reading control • Data use transparency • Usable privacy • NonretenBon of data
2. Data Portability
The user is able to obtain their data and to transfer it to, or subsBtute data stored on, a compaBble plaUorm.
Associated concepts: • Free data access • Open formats • PlaUorm independence • Free deleBon
3. CreaBve Control
The user is able to modify their data within the soCware plaUorm being used, and to control who else can do so.
Associated concepts: • Originator-‐discreBonary ediBng control • Authorial copyright support • Reciprocal data sharing
4. SoCware Freedom
The user is able to modify code in the soCware plaUorm being used, subject to rights of other users to control their own experience of the plaUorm.
Associated concepts: • Open Source code • Reciprocal code openness • User modifiable plaUorm
5. ParBcipatory Design
The design of the plaUorm is produced by all of its users.
Associated concepts: • User-‐centered design • User input to design • User-‐generated design • Customizable design
6. User Self-‐Governance
The operaBon of the plaUorm is governed by all of its users.
Associated concepts: • ParBcipatory policy making • ParBcipatory implementaBon • ParBcipatory adjudicaBon
7. Universal Network Access Every person is legally and pracBcally able, to the greatest extent possible, to access the Internet, and it is available everywhere in a form adequate for both retrieving and posBng data. Associated concepts: • Right to connect • Universal digital literacy • No-‐ or low-‐cost service • Omnipresent service • Accessibility
8. Freedom of InformaBon Every person is legally and pracBcally ab l e t o p roduce and r e ce i v e informaBon in the way that they want, to the maximal extent consistent with the rights of others.
Associated concepts: • Right to privacy • Right to anonymous speech • Freedom from censorship • Open Access to all publicly funded data • DemocraBcally controlled security • Right to be forgoJen
9. Net Neutrality All providers of Internet connecBons and services are legally and pracBcally required to treat data equally as it is transmiJed through the infrastructure they control. Associated concepts: • Source neutrality • Format neutrality • Content neutrality • End-‐user neutrality
10. PluralisBc Open Infrastructure
Everyone has access to mulBple independent but interoperaBng soCware plaUorms as opBons for their data.
Associated concepts: • MulBplicity of plaUorms • Decentralized control • Transparent control
Ten Principles in Three Clusters User Data Freedoms 1. Privacy Control 2. Data Portability 3. CreaBve Control
So8ware Pla:orm Freedoms 4. SoCware Freedom 5. ParBcipatory Design 6. User Self-‐Governance
Public Network Freedoms 7. Universal Network Access 8. Freedom of InformaBon 9. Net Neutrality 10. PluralisBc Open Infrastructure
Survey Research QuesBons RQ1: Do users (who are knowledgeable enough to understand the ten principles) aJach stable values to the principles? RQ2: Is there a common value that correlates with the values users aJach to all of the principles? RQ3: Do the values that users aJach to the ten principles differ qualitaBvely from each other and/or fall into clusters?
ParBcipants and CondiBons
780 P’s recruited on MTurk from the U.S., filtered by approval rate 2 groups, randomly assigned: (a) Broad ra@ng set: all ten of the main principles only (b) Narrow ra@ng set: one of the main principles and all of its associated concepts (10 subgroups)
Procedure
P’s shown only the descripBon of each principle or concept, not its name • e.g. “The user is able to know and to control who is able to access their data.” [Principle 1: Privacy control]
• e.g. “Policies ensure that all users have mulCple soDware plaEorms to choose from as environments for their data.” [Concept 10a: MulBplicity of plaUorms]
P’s asked for each principle or concept, “How important is this to you? (0=not at all, 10=extremely)”
1-‐Privacy Control
2-‐Data Portability
3-‐CreaBve Control
4-‐SoCware Freedom
5-‐ParBcipatory Design
6-‐User Self-‐Governance
7-‐Universal Network Acces
8-‐Freedom of InformaBon
9-‐Net Neutrality
10-‐PluralisBc Open Infrastructure
0 1 2 3 4 5 6 7 8 9 10
1
2
3
4
5
6
7
8
9
10
Aggregate Importance Ra@ngs
10 Main Principles
RQ1: Are importance raBngs stable for the ten main principles?
Two contexts for comparison: (a) Broad raBng set: all ten of the main principles only => within-‐subjects raBngs (b) Narrow raBng set: one of the main principles and all of its associated concepts (10 subgroups) => between-‐subjects raBngs
Within-‐ Vs. Between-‐Subjects Importance Ra@ngs CorrelaBon = 0.98 => Stable raBngs for the ten main principles
0
1
2
3
4
5
6
7
8
9
10
0 1 2 3 4 5 6 7 8 9 10
Within-‐Subjects
Importance Ra@ngs
Between-‐Subjects Importance Ra@ngs
RQ2: Is there a single underlying value that predicts all importance raBngs?
Broad raBng set (within-‐subjects raBngs of 10 principles) yields intrapersonal correlaBons across the principles Two sub-‐quesBons: • Are all correlaBons posiBve? • All all correlaBons significant?
Comparing Importance RaBngs (Ten Main Principles)
43 out of 45 correlaBons are posiBve 29 out of 45 correlaBons are significant (all posiBve) => Not quite one underlying value
RQ3: Do the principles’ raBngs differ qualitaBvely/fall into clusters?
RaBngs for the lowest rated principles (4-‐SoCware Freedom and 5-‐ParBcipatory Design) were significantly and posiBvely correlated (.36) But… • only 3 out of 14 correlaBons between principles 4 or 5 and other principles were significant
• and 25 out of the remaining 30 correlaBons were signficant
(p=.0001 by a Fisher exact test)
1-‐Privacy Control
2-‐Data Portability
3-‐CreaBve Control
4-‐SoCware Freedom
5-‐ParBcipatory Design
6-‐User Self-‐Governance
7-‐Universal Network Acces
8-‐Freedom of InformaBon
9-‐Net Neutrality
10-‐PluralisBc Open Infrastructure
0 1 2 3 4 5 6 7 8 9 10
1
2
3
4
5
6
7
8
9
10
Aggregate Importance Ra@ngs
10 Main Principles
Survey Research QuesBons: Summary
RQ1: Do users (who are knowledgeable enough to understand the ten principles) aJach stable values to the principles?
-‐> Yes, across broad and narrow ra@ng sets RQ2: Is there a common value that correlates with the values users aJach to all of the principles?
-‐> Not quite, but most are posi@vely correlated RQ3: Do the values that users aJach to the ten principles differ qualitaBvely from each other and/or fall into clusters?
-‐> Yes, So8ware Pla:orm Freedoms (principles 4, 5, & 6) rated lower, less correlated with other principles
Four Users’ Rights Frameworks
• Rights and ResponsibiliCes of Electronic Learners (RREL, 1994) [Frank Connolly]
• A Bill of Rights for Users of the Social Web (BRUSW, 2007) [Joseph Smarr et al.]
• Marco Civil da Internet (MCdI, 2014) [Brazilian Government]
• NETMundial DraD Outcome Document [NDOD, 2014) [“MulBstakeholder”]
Conclusion #1 Privacy Data Portability CreaBve Commons Free SoCware Open Source ParBcipatory Design Wikipedia-‐Style Democracy Right to Connect Accessibility Freedom of Expression Open Access Right to Be ForgoJen Net Neutrality Open Protocols
All of these should be seen as possible elements of a more general Users’ Rights Framework
Conclusion #2
An analysis framework for users’ rights can be used to compare policy frameworks, to see what they have in common as well as missing elements.
Conclusion #3
Internet users who are knowledgeable enough to understand users’ rights principles have stable attudes toward them, and tend to value them highly.
Conclusion #4
Despite the efforts of those who have advocated for them, what I have called “SoCware PlaUorm Freedoms” (especially soCware freedom, parBcipatory design) appear to be viewed as dispensable relaBve to other digital rights and freedoms, even by most sophisBcated users such as screened MTurk workers. (Is this attude jusBfied?)
top related