You are viewing a read-only archive of the Blogs.Harvard network. Learn more.

06/16/05 Meeting Notes

These notes are a best effort.

Blog your corrections and commentary.

Attendees:

  • WK: Wendy Koslow
  • EG: Erica George
  • DF: Deborah Elizabeth Finn
  • BK: Beth Kanter
  • guyinsuit
  • LB: Lynne Baehr
  • PF: Paul Frankenstein
  • RF: Randy Fenstermacher
  • BS: Brett Stilwell
  • UG: Urs Gasser
  • LW: Lisa Williams
  • MW: Mal Watlington
  • CB: Carl Blesius
  • j: j

Agenda:

  • Berkman Center fellow Urs Gasser on information quality:

    The question of credibility of information posted on blogs or news sites has become a hot issue. Urs argues that this is only the tip of the iceberg or, more precisely, the most visible expression of a much broader and more fundamental question that accompanies the transition from analog/offline to digital/online media, i.e., the question of information quality. Tonight, Urs will give a 10-minute-overview of his forthcoming paper that explores this issue. The focus is (1) on the effects of the Internet on information quality and (2) on possible regulatory responses (in the broadest sense of the term ‘regulation’.)

    • UG: Blogs as a medium, has shifted way people communicate and also access information. Not like an offline library for example.
    • UG: We’re much more interactive – mashups, fanfiction, etc.
    • UG: blogs illustrate that these shifts have both + and – impact on quality of information (I missed 2 shifts – EG). Example: the Bush national guard/Dan Rather fiasco. Bloggers found that the documents presented by CBS were probably forged. So there, blogs enhanced the quality of available info.
    • UG: Lots of problems with inaccurate info, especially for health information, alternative disease treatments, etc.
    • UG: So then what *is* quality information? – No simple definition.
    • UG: twofold problem. Information itself has many meanings. I talk about information as semantic. quality is meaning & message, content, not bits & bytes.
    • UG: Also, what do we mean by quality? Lots of research, lots of different sets of criteria for “quality” info. Accuracy, timeliness, etc pop up a lot.
    • 4 categories of information quality criteria:
      • functional/factual/tangible
      • cognitive – is info complete? very dependent on the knowledge of individual user
      • ethical – is it a fair representation, ethically appropriate description, etc.
      • aesthetic – is the poem good? etc.
    • UG: but these categories may be irrelevant in context of information quality (I’ll call it IQ – EG) regulation.
    • UG: ways to regulate: laws, markets (by price), social norms, technology
    • UG: What kind of IQ we’re dealing with affects how it can be regulated. Much easier to regulate on function than on aesthetics or ethics.
      • multidimensional
      • relational – it’s human beings who make IQ decisions (subjective)
      • contextual – situations determine informational needs
    • What ae the issues for IQ on the net?
      • examples: creation category:
      • anyone can add information to a medical database.
      • edit wars on Wikipedia.
      • disagreement re accuracy of feedback on eBay
      • examples: distribution
      • distribution is global. information produced in one local context is accessible in many other contexts. different criteria for evaluation.
      • French yahoo site search results that showed nazi memorabilia, which violated French law. Different approaches to judging what kind of speech is accessible.
      • examples: access
      • accountant in CA sued Google for inaccurate search results when his name typed in
      • jew search showed an antisemitic site first, leading to grassroots googlebomb to make wikipedia article the top result
      • SR: Q: if IQ encompasses all these things, what if any are general conclusions we can make?
      • UG: Getting to that. These examples lead to the initial conclusion: This is big, manyfaced, too faceted to come to a general conclusion. But also, there are internet-specific quality concerns (ie Wikipedia editing) that don’t happen in traditional publishing media like books (even authors vs editors) or newspapers. And third, that the question isn;t just theoretical, but that it leads to real issues that need to be resolved.
      • UG: And if this leads us to saying there are unresolved issues, the next obvious question is, what are possible responses?
      • One is that there ought not be any quality regulation on the net, nobody even has the power to regulate it, it’s too big, and more so, nobody ought to. It’s not necessary. Marketplace of ideas. The good ideas will win out, the truth will win over false. It’s a good argument, but I don’t buy it, entirely.
      • DF: That’s predicated ona certain education level. Like democracy predicated on a citizenry with grasp of civics. The marketplace of ideas depends on media literacy.
      • RF: but embedded assumption is that education is intrinsically higher-value information that information not gained through education
        <li.DF: not all edu is in classroom. But there is such thing as being literate in public issues, in natrue of the internet etc

      • RF: sometimes wikipedia is better quality than academic literature.
      • UG: but wikipedia is not laissez faire. It’s very strongly controlled by social norms. Moderated by editors, albeit decentralized. Guideline of neutral viewpoint. rules of the community.
      • UG: I haven;t seen this marketplace of ideas actually work in practice without being moderated somewhat in actual practice
        <li.SR: i don’t even see how this laissez faire thing could even be possible. Information is communication. Piece of communication can’t be isolated from its producers and consumers. Social structure then leads to power structure leads to possibility of social regulation.

      • UG: even in a “marketplace of ideas” there are quality assessments. They’re based on something. Policy perspective, the idea is that to answer the problem, add more info. Just increase the number of speakers and solve all the quality issues
      • MW: that the solution to pollution is dilution 🙂
      • MW: in other marketplaces that use quality as a criterion, quality just means “conformance with requirements”. So it’s by definitoon contextual on the needs of the end user. But i hear discussion on IQ always going to the absolutes.
      • UG: Ok, the other things beyond marketplace of ideas. I disagree that simply more info is the only needed answer to all IQ problems
      • 2nd response is an information order model. Such as China filtering everyting, choosing what is “good” info what is “bad.”
      • UG: But this could also be implemented democratically. A govt could say that IQ is so crucial, it requires laws. Standards. Ex, that Yahoo case in France. Govt itself becomes provider of quality info: BBC model. This model assumes one centralized force, an agency, govt body, etc.
      • UG: 3rd model, is my favorite
      • is decentralized model, narrowly tailored to needs of smaller groups or types of uses. philosophy that IQ is indeed important but there are so many value questions in play, fundamental norms such as autonomy that make it hard to regulate info. Not just for governments to regulate, but other individuals – nobody wants to be told what is good or not good for them, they want to decide themselves.
      • This model does use market mechanisms. Think online reputation systems, allexperts.com, trust marks on health sites…
      • RF: this is meta: information about information. the classical market is things not info. then there’s the market for info. then the market for info about info. with goods everyone can see the good, see what it costs, etc. Now information is a commodity, and not only that, information about information is a commodity.
      • LW: i can’t remmber the last time I (directly) paid for a piece of information on the internet. We’re inverting the market. Normally for goods the info is free, the good costs money. Now you just get snippets of info because the info is the good.
      • LW: Linking is like price now. Linking shows the relative value of something. But you can’t pay the rent (yet) with links.
      • UG: A nobel-prize paper on markets of lemon cars: you don’t know the quality of the car. Ackerloff. Asymmetric information causes market failure. Then a few years later Spencer (?) shows that prices aren’t the only indicator. Other market solutions signal quality. One is reputation. Trust marks, education (people use the education to signal quality of a potential employee).
      • UG: problem is to figure out what is the profile of each approach to regulation, how does it fit, what sets of problems can this approach solve. (When are social norms enough to solve problem? When do you need technology? Laws?)
      • UG: This 3rd approach to the problem of IQ I like. Acknowledges shift from a world focused on pure growth, pure “more”, to a world more oriented to quality. But I prefer that quality be regulated as non-centrally as possible (EG: I’m not fully sure this last is what he said)
      • MW: how do you track progress when quality is harder to measure than growth (EG: I think this is what he asked)
        <li.UG: autonomy, diversity, are very valuable. also, level of harm/non-harm. We have a war that was started on bad information. Information quality affects lives. Some spaces are better for regulation than others, if want to preserve open and free space.

      • LW: IMHO we regulate action more than information. we have a right to bad information more than the right to cause physical harm for example. What kinds of info have material impact on my life. A website on wild mushrooms could protect or mislead me in terms of what’s poisonous. But for many kinds of info it’s ahrd to argue harm unless it’s based on or leads to an action.
      • RF: there are some political groups that do feel that some information is itself harmful. ie religious groups
      • WK: ie gag rules re talking about abortion if funded by US govt overseas. Where to draw the line?
      • SR: sensitive financial data – requires certifications, etc. We apply stringent requirements to sensitive transactions that we don’t apply to, say, movie reviews. We accept laws that apply to certain transactions, in others we don’t need the law.
      • MW: law is a trailing edge function.
      • LW: we make laws after someone gets hurt. that’s part of the american experiment. we tend not to presumptively/pre-emptively regulate information. we tend not to speculate that someone might be hurt
      • RF: but the current republican congress *is* doing that
      • UG: There are procedures so we can agree on laws, or express our disagreement. But it’s not just laws. A code of ethics can simply be generally socially agrees on. There are several drafted blog codes of ethics. Some use them some don’t. Some people have a blog transparency statement. That can have a regulating function. It provides you infroamtion on which to evaluate the information from a person. It’s very much not just law.
      • UG: re harm, it’s not just physical. Harmful speech may not harm physically, at least not directly. But we rate movies for a reason. A piece of information can impact a state of mind. Can impact behavior. Even laws acknowledge this. Esp. re health info, health privacy (HIPPA), etc.
      • UG: to keep cyberspace free and open, down the road we will need to deal with the question of quality. there are limits to growth. more is not always better. there are costs to unrestrained growth. we must improve as well as grow. My conclusion: more *plus* better.
      • UG: and now, I’ve got to run, sorry! 🙂
  • Changes in next week’s agenda: iLaw-related blog training with Rebecca MacKinnon PLUS the scheduled Wiki Wonderland with our beloved SJ Klein!
    • WK: Last year, RMack and Dave Winer did a blog tutorial during iLaw. We wanna do one again this year. Sammy is cool with combining this with his wiki presentation. We anticipate maybe 10 iLaw participants. And there’ll be pizza!
  • DF: cool thing: Beth and I will be blogging at iLaw! yay!
  • LW: Pending item: Now that we’re halfway through the previously scheduled stuff we may want to start thinking ahead.
  • Poss changes to schedule: Bill Ives for June 30, Nonprofit blogs for July 7. To be confirmed.
  • LB: Random: Livejournal now has tags!
  • EG: w0000t!!!
  • Well wishes for Gregory
  • What else?
  • Eat: Uno’s
    • Last week, Newbury St.
Log in