Joho the Blogtrust Archives - Joho the Blog

January 24, 2012

[2b2k] Trails of Trust

Panagiotis Takis Metaxas (at the Berkman Center) and Eni Mustafaraj have written a paper called “trails of Trustworthiness in Real-Time Streams” [pdf] about how to support critical thinking about social networking conversations, while maintaining privacy. From the abstract:

When confronted with information that requires fast ac- tion, our system will enable its educated users to evaluate its provenance, its credibility and the independence of the multi- ple sources that may provide this information.

They say the only real hope is to solve the problem within closed streams that provide membership functions because there “it is possible to determine the a priori trustworthiness of a message received,” by evaluating the credibility of users on particular topics. They believe this can be done by watching the actions of users. For example, “In general, the more often a user re-posts messages from a sender, the more trusted the sender becomes.” And: “A message that has been sent by different, independent users has more trustworthiness than one that has been initiated by a single user.”

There’s much more in their paper…

Comments Off on [2b2k] Trails of Trust

May 31, 2011

[berkman] Miriam Meckel on communicating trustworthiness

Miriam Meckel is giving a Berkman lunchtime talk on “Communicating Trustworthiness: Drivers of Online Trust.” She will present research she has been doing at U. of St. Gallen along with government and some businesses. She’s investigating how trustworthiness is communicated, and what the initial drivers and cues are. She’s going to look at the changing conditions, some basic ideas on initial trust formation, then her study, and then the cues identified in initial trust formation.

NOTE: Live-blogging. Getting things wrong. Missing points. Omitting key information. Introducing artificial choppiness. Over-emphasizing small matters. Paraphrasing badly. Not running a spellpchecker. Mangling other people’s ideas and words. You are warned, people.

Users’ willingness to provide data about themselves depends on some sort of trust. There’s been much discussion about this, including the importance of preserving privacy. For example, Facebook restricted info about users to the user’s schoolmates, but over time it has become far more permissive. (See here for article and diagram.) Both Facebook and Google have said things indicating they believe our attitude towards privacy has changed. E.g., Eric Schmidt: “If you have something you don;’t want anyone to know, maybe you shouldn’t be doing it.”

What is trust? Miriam provides three definitions, each of which involves vulnerability. When we trust someone online, we understand we are making ourselves vulnerable in various ways. Trust enables us to manage the complexity of our social world, reduces the cost of negotiations, facilitates the adoption of new technologies, and reduces perceived risks. Initial trust is especially important since you can lose a customer forever by screwing it up.

In her study, they interviewed 23 online businesses active in Germany. They also did in-depth interviews with 43 users representing a range of socio-demographic sections. From this they derived a model of nine trust drivers: Reciprocity and exchange, proactive communications, user control, brand, third party endorsements, design of user interface, offline presence, technology, and customer service. Then they did a quantitative survey of German users to find out how the different factors are applied to diff business models. 11,000 users from across socio-demographic segments were invited to participate in an online survey for a little money, with a response rate of 12%.

She goes through the nine components of trust:

Reciprocity explains about 35% of the variance of the data. The mutual relationship of customer and organization is crucial. Users are willing to hand over personal info, if they know what they gain from it, if it is transparent, and if both sides benefit.

Brand and reputation: A well-known, establish brand helps build trust. If it has a large customer base, it is perceived as more reputable. You also need a professional feel for the web site.

User control: Users want to know who will have access to their data. Ask permission.

Proactive communications. [missed it. sorry]

Customer service: Multiple ways to reach a person. Provision of different payment methods.

Offline presence: The existence of physical stores helps build trust. Show a photo on your web of your brick and mortar store.

Technological reliability. This can be an issue when your service involves third parties.

Third party endorsements: Not only are seals of approval helpful, but so is having a high search ranking for third parties.

The next step is how these factors are differentiated by B2C business models. Only four factors turn out to be relevant across the board: reciprocity, third party endorsements, user control, and technological reliability. Miriam looks at online shopping, online banking, etc., to see which factors are relevant [but I can’t keep up with my typing. sorry] It seems that for building trust-based relationships with potential clients, you have to have fair communication, mutual benefits signaled, privacy and security policies being displayed, clear T&C’s, engage in issues mgt to follow what is being said about you on the Web, explain the business model and data needs, explain the flow of info to third parties and what the policies are, communicate third party endorsements, engage, peer groups and communities, bring in your offline reputation, and consistently apply corporate design with good design.

Some implications: 1. Users are more willing to trust large, well-established and popular online services. High search-engine ranking can be considered a trust measure.

2. The visual appearance counts.

3. Reciprocity is very relevant in all online transactions.

4. Offline presence, technical reliability and customer service have barely been researched yet.

5. Organizations need a strategic approach for communicating trust to their stakeholders. Take an integrated communication approach, including issues and risk management. They should communicate proactively, in a conversational tone, and transparently.

What not to do? Don’t be Facebook asking Burson Marsteller to find bloggers who would attack Google.

Q: [me] Customer ratings? And the presence of customized trust mechanisms as at eBay and Amazon?
A: Third party endorsements include customer ratings. But we surveyed a represented group of German users, not ones as sophisticated as others, plus there may be cultural differences.

Q: Is your work ethical? An evil company can read your work and figure out how to appear to be ethical. Also, Germany has stronger regulatory protection, which change the signals
A: Yes, companies could fake being trustworthy.

Q: How about SSL?
A: The awareness of the need for SSL is frighteningly low.
Q: Anyone care about authentication certs?
A: [general laughter]

Q: Would the same trust drivers be applicable in news sites and other such non-sales sites?
A: Just in parts. [I can’t read her table because I’m too far back :( ]

When you look at the erosion of privacy caused just be clicking on opt-ins, you get to a point where you just accept whatever the terms are and however they’re changed.
A: People don’t read privacy policies when they’re changed.
Q: Maybe it’s that when the privacy policy looks impressive, it’s a signal of the quality of the service. But, the key question is whether people who say privacy policies are a trust driver actually read them.
Q: There’s some survey support that most people mean think that a privacy policy is there to protect your privacy. [general laughter]

Q: [me] To me, third party endorsement means you get celebrities or associations or other companies to endorse you. But the real trust-driver for me frequently is peer endorsement. Maybe it would be worthwhile to separate those two, especially since businesses need to do different things to gather other companies’ endorsements and to gain a good reputation among peers.
A: Yes, that would be interesting.

Q: What supersedes trust?
A: There may be the desire to be “in” as opposed to “out.”
Q: There are network effects are. How many of friends have to join before I throw away my privacy cBerkman

Comments Off on [berkman] Miriam Meckel on communicating trustworthiness

February 24, 2009

[berkman] Pippa Norris on cultural convergence

Pippa Norris of the Harvard Kennedy School is giving a lunchtime Berkman talk titled “Cultural Convergence: The Impact on National Identities and Trust in Outsiders.” [Note: I’m live-blogging, hence making mistakes, missing stuff, misunderstanding other stuff, typing badly. This is an inaccurate, incomplete record of her talk.]

What might be the impact of cosmopolitan communications, she asks? Her thesis is that there are many firewalls that block global information flows. She will argue that the news media has an impact through cosmopolitan communications, and will look at the implications for public policies. It makes people slightly less nationalistic. [Note: She talks fast. Bad for live bloggers, but good for listeners.] (This is from her book, available free on her Web site.)

Globalization is the starting part. It’s about more than trade; it’s also social and political. Cosmopolitan communications = “the way we learn about, and interact with, people and places beyond the borders of our nation-state.” Cosmocomms have been expanding. But, so what, she asks. In the 1970s, this was seen as cultural imperialism. In the 1990s, it was thought of as Coca-colonization. In the 2000s, we’re still seeing cultural protectionism.

Pippa will focus on audio-visual publishing. Western countries remain dominant. In fact, the gap has widened. There are four views in the literature: 1. There’s a convergence around US exports. 2. There’s a polarization of national cultures. 3. There’s a fusion of national cultures. 4. Pippa’s firewall model.

The firewall model says that there are barriers: 1. Trade barriers; 2. Internal barriers to free press; 3. Poverty; 4. Learning barriers that make it harder to acquire values and attitudes.

She discusses three levels: individual, national, and cross-level. For the national, she talks about the “cosmopolitanism index” she’s devised. She’s surveyed 90 countries using a set of survey questions. At the bottom are the poorest countries with the poorest press freedom. At the top, Luxembourg, Switzerland, Norway, Denmark. It goes from 1972-2004.

That’s the framework. She confines her discussion of the results to effects on national identity and trust. Roughly, the more cosmpolitan, the less national identity and higher trust in outsiders. In terms of trusting outsiders, Norway takes the cake. US and Sweden are also high. (The Netherlands are high on the cosmo scale but only in the middle on the trust scale. Also, Germany and Spain.) But because there’s no control group — maybe trust maps to education? — you have to do multilevel regression. She uses age, gender, income, education, and news media use, and finds that trust correlates with news media use.

Her conclusions: Use of the news media “is positively related to more trust in outsiders” (different countries and religions) and “is related to weaker feelings of nationalism.” “I regard that as positive results.” But there are some qualifications: 1. Many other factors create trust in outsiders. 2. This study looks at the impact of news media, but not the impact of entertainment media. 3. There may be self-selection bias or interaction effects.

Policy implications? Is the globalization of news media a threat to national diversity? See

She concludes by asking what we know about how we measure flows of info from one country to another, over time, say from 1995?

Q: [ethanz] There are some familiar data sets. E.g., Alexa, although because it’s opt-in, it’s not perfect. There’s also Google Search Insights that tracks searches. In most countries, “news” is almost always one of the very top searches. A question: How might your analysis integrate with national-level studies. E.g., a study that showed that as cable TV was introduced to communities in India, you got an increase in empowerment equivalent to 4 years of education. [I probably got this substantially wrong.]

Q: There are categories of trust…
A: We use the World Values survey that includes over a quarter million people. Is trust in Nigeria and Sweden the same? There are many categories indeed. But when you see a strong pattern emerge, as we have, then we should assume something is happening in the data.

Q: A speaker from Microsoft was at Berkman recently, talking about the issues importing and exporting data on the Net across national boundaries. What sorts of measures have you been using?
A: The obvious ones. Internet access. Location of hosts. And some articles that have looked at search terms.

Q: I’m from Poland: High cosmo, low trust. In the US, we rent movies instead of watching TV. But rental stores don’t know about a particular movie highly famous in Europe. My question is about the global dimension of local issues.
A: Poland and much of Central Europe have suddenly become much more open and have found greater value change than in countries open for a long time. You should see greater variation within such countries, e.g., by age.

Q: [smacleod] Pippa asked for ideas about media flows, with some positivist assumptions about the ability of globalization and media studies to be objective. Has she read Appadurai’s “Modernity at Large: Cultural Dimensions of Globalization,” which emphasizes the mutability of flows. I wonder how Pippa might engage on-the-ground research and how such ethnographic research might recast her methodological assumptions. Extensive anthropological field work has been done in all the countries she’s mentioned, which engages and historicizes the legacy of colonialism.
A: I am a positivist. But I also like dealing with cases. There’s rich work being done in communications and other fields, but there’s also good division of labor. You need both. Some people like to fly over a country and others like to walk through it, and you get value from both.

Q: The US has more cultural exports than imports, while most seem to have about equal amounts. How does this play into cosmo?
A: America also imports a lot. America doesn’t have to import a lot because it’s got so much.

Q: Tribal populations in America have a tight tie to geography. Where’s UNESCO is generating the data to look at other regions than nation states?
A: The data generally depend on national statistical offices. UNESCO depends on those; it has no data generation capability itself.

Q: [hal] Google Ad-Planner lets you download a list of the 500 most visited sites for many countries. It has unique visitor numbers.
A: So I could see how many people go into Norway and how many go out.

This gives us a way to focus on globalization of media by focusing on the people, not on the media, Pippa concludes, reminding us that the chapters are up on her site. [Tags: ]

1 Comment »