Truck and Barter

Social Networks and Country-Wide Content Filtering

Apropos of nothing really, via Slashdot I see that a fascinating new product to route around country-wide internet filters, called Psiphon, has been released for public download.

NB: Amusingly, the actual Psiphon site is actually blocked by a number of corporate filtering programs. To learn about a few details, take a read through either this article, or the Wikipedia entry on it.

The idea really isn't new, as pointed out by the Slashdot posting, and Psiphon is just another in a line of proxy and filter avoidance technology. Just another step in the arms race between those looking to keep tight control over internet access and those who want to get around it. Now, if only there were a good, analytical way to study the process of interactions in arms races.

Kevin, any thoughts?

Learn from the England

In earlier post I commented that Brookings had suggested that State Department should follow DFID’s lead in the development aid. Now Posner suggests, “We Need Our Own MI5”;

“Intelligence succeeded in part because of the work of MI5, England's domestic intelligence agency. We do not have a counterpart to MI5. This is a serious gap in our defenses. Primary responsibility for national security intelligence has been given to the FBI. The bureau is a criminal investigation agency. Its orientation is toward arrest and prosecution rather than toward the patient gathering of intelligence with a view to understanding and penetrating a terrorist network….

The bureau's tendency, consistent with its culture of arrest and prosecution, is to continue an investigation into a terrorist plot just long enough to obtain enough evidence to arrest and prosecute a respectable number of plotters. The British tend to wait and watch longer so that they can learn more before moving against plotters.

The FBI's approach means that small fry are easily caught but that any big shots who might have been associated with them quickly scatter. The arrests and prosecutions warn terrorists concerning the methods and information of the FBI. Bureaucratic risk aversion also plays a part; prompt arrests ensure that members of the group won't escape the FBI's grasp and commit terrorist attacks. But without some risk-taking, the prospect of defeating terrorism is slight.

MI5, in contrast to the FBI (and to Scotland Yard's Special Branch, with which MI5 works), has no arrest powers and no responsibilities for criminal investigation, and it has none of the institutional hang-ups that go with such responsibilities. Had the British authorities proceeded in the FBI way -- rather than continuing the investigation until virtually the last minute, which enabled them to roll up (with Pakistan's help) more than 40 plotters -- most of the conspirators might still be at large, and the exact nature and danger of the plot might not have been discovered. We need our own MI5, not to supplant but to supplement the FBI…”

More at their weblog.

‘Free the Curricula, Free the Textbooks’- A Wikiality?

Two articles on the history of Wikipedia via Marginal Revolution; The Hive and Can Wikipedia conquer expertise?

Another explanation from The Economist;

“This success has made Wikipedia the most famous example of a wider wiki phenomenon. Wikis are web pages that allow anybody who is allowed to log into them to change them. In Wikipedia's case, that happens to be anybody at all. The word “wiki” comes from the Hawaiian word for “quick”, but also stands for “what I know is...”. Wikis are thus the purest form of participatory creativity and intellectual sharing, and represent “a socialisation of expertise”, as David Weinberger, who is currently writing a book on collaborative intelligence, puts it.

Among the new media, wikis are the perfect complement to blogs. Whereas blogs contain the unedited, opinionated voice of one person, wikis explicitly and literally allow groups of people to get on the proverbial “same page”. This is the main reason for the failure of a Los Angeles Times experiment with wikitorials, described in the previous article. Wikis are good at summarising debates, but they are ill-suited for biased opinion.”

Here’s Colbert’s attempt at explaining the Wikipedia; see also this video.

The major innovation I’m looking forward is when the Wikibooks gets a real take off- I don’t think it’s wikiality!

Related Links;

Best coverage of the Wikipedia amongst the blogs is at at Ross Mayfield’s Weblog.

Wikimania 2006: Opening Session with Jimmy Wales

Ten - or maybe a dozen - Things that Will Be Free

Wikimedia Foundation

Internet encyclopaedias go head to head; Jimmy Wales' Wikipedia comes close to Britannica in terms of the accuracy of its science entries, a Nature investigation finds.

The Faith-Based Encyclopedia

A couple of interesting posts by John Quiggin; Wikipedia and Sausages, Wikipedia-economics-category-project, When co-operation trumps competition

The Wealth of Networks

Multimedia;
Digital Maoism; here is the transcript.
Interviews from a Survey of New Media in The Economist; Andreas Kluth, technology correspondent
David Sifry, Founder and CEO, Technorati
Chris Anderson, Editor-in-Chief of Wired
Jerry Michalski, founder and president of Sociate
Paul Saffo, Director, Institute for the future

Researching with Wikipedia- introductory videos

Brion Vibber has worked on MediaWiki and Wikipedia's; an engineers view

Somebody Not Happy with Wikipedia

Did you see the broken light?

“Three decades ago, a group of students were shown a short movie in which two cars were in an accident. Then the students were divided into two groups where the first group was asked "Did you see the broken light?" and the second was asked "Did you see a broken light?" Switching one single word, the or a, in the otherwise identical question changed responses by an astonishing 31 percent.

A body of literature has shown that there are many ways to influence respondents, too often too subtle to be recognized. You can probably guess that using the word "financial incentives" or "subsidies" will elicit different results. But would you think that the order in which different alternatives are presented to the respondent might influence his or her response? Probably not, but in reality it does.

Irrespective of how the question is worded, survey methods that could influence the data collected, such as using or not a public official as interviewer or reading the questions to the respondents instead of showing them written questions are known as survey fixed effects. Not taking such effects into account can bias the results, says Iarossi.”

- A review of the book, The Power of Survey Design: A User's Guide for Managing Surveys, Interpreting Results, and Influencing Respondents by by Giuseppe Iarossi

The book is a must read for anyone interested in anything to do with surveys.

Some other book reviews in the latest F&D.

Better Living Through Better Design

In a highly cluttered visual field, people have a harder time picking out a target of interest, and often choose wrongly. Not a surprise, that. More interestingly, however, is that people tend to have high confidence that they are correct.

One might intuitively expect that as background noise created by distracters and errors increase, confidence in one’s decision plummets. But in a new study published in PLoS Biology, Stefano Baldassi, Nicola Megna, and David Burr show that just the opposite happens. When they asked observers to search for a tilted target embedded in vertical distracters and estimate the target’s tilt, the observers often overestimated the magnitude of the tilt--and did so with a high degree of confidence in their decision.

The authors used signal detection theory to make quantitative predictions about the probability that an observer will detect a target under cluttered conditions. SDT assumes the brain represents each element in a visual search display as an independent variable with its own noise. It also assumes that when the observer isn’t sure which stimulus is the target, she monitors all stimuli, and performance suffers. Thus, increasing the number of distracters (trying to find your friend on a busy street or a document on a messy desk) increases the background noise of the visual system’s representation while reducing the accuracy and reaction time of performing the task.

As a general condition, I wonder what role this might play in individual decision-making over a vast range of choices. Specifically, I got to thinking about the contentions some make about having "too much choice." (See the Barry Schwartz quote in Postrel's post.) The claim that too much choice makes people worse off, ably dealt with in Postrel's full article, leads to the conclusion that choice ought to be limited. But what if the claims of dissatisfaction with abundance are simply picking up on a different problem?

From the Innovations Report article:

The authors explain that while their study focused on "simple perceptual decisions about a single stimulus attribute," the same type of processes may also apply to complex cognitive tasks involving problem solving and memory. If people find themselves confronted with multiple events in a chaotic, confusing environment, they may decide about some aspect of the situation and be totally wrong even though they have full confidence in their decision. The consequences of such a phenomenon could be relatively trivial, explaining why professional athletes often end up wasting their time arguing questionable calls with an official.

Sounds to me like being presented with complex visual fields is a bit disrupting to a lot of people. So much so that it results in frustration. While it might not be described as "chaotic" in terms of movement, the whole aisle of toothpaste options noted by Karrie Jacobs observes in Postrel's article, as well as the vast array of items in shopping malls, car dealerships, convenience stores, the newspaper ad pages...all certainly presents a cluttered visual pattern, what with all of the conflicting colors, shapes, lettering, and advertising thrown in by the store itself. Perhaps those negatives that Shwartz says pile up are a result of the displeasing sensation from the visual information overload, and not necessarily a dislike for the amount of choice.

Maybe the problem is just that we have really bad graphic design.

Journal Ranking By Google?

A new method for ranking science journals is being proposed. What's it based on? Google, of course.

The most popular index of a journal's status is the ISI Impact Factor (IF), produced by Thomson Scientific. It counts the total number of citations a journal's papers receive, and divides it by the number of papers the journal publishes. But the rise of online journals, coupled with sophisticated search engines that permit rankings of web resources, is triggering a wave of other measures. Last year, for example, physicist Jorge Hirsch of the University of California, San Diego, proposed a metric called the h-index for assessing the quality of researchers' publications (see Nature 436, 900; 2005).

Now Johan Bollen and his colleagues at the Research Library of Los Alamos National Laboratory in New Mexico are focusing on Google's PageRank (PR) algorithm. The algorithm provides a kind of peer assessment of the value of a web page, by counting not just the number of pages linking to it, but also the number of pages pointing to those links, and so on. So a link from a popular page is given a higher weighting than one from an unpopular page.

Only hearing second-hand information about the way tenure review goes for academics, I was under the impression that the relative "importance" of the journals in which one might publish also has a big impact on tenure decisions. Ranking systems then seem to be deeply involved in the way research occurs if it impacts who gets financial support either throught grants or university support.

My question is whether the "PageRank" metric might be highly vulnerable to information cascades and manipulation in much the same way as Google's process for ranking sites.

Catastrophic Failure

This is only marginally relevant to the general discussion around here, but it does highlight a point I've raised in the past.

Looks like it's a lot easier to crack the RFID on a passport than one might hope.

Dutch TV programme Nieuwslicht (Newslight) is claiming that the security of the Dutch biometric passport has already been cracked. As the programme reports here, the passport was read remotely and then the security cracked using flaws built into the system, whereupon all of the biometric data could be read.

The crack is attributed to Delft smartcard security specialist Riscure, which here explains that an attack can be executed from around 10 metres and the security broken, revealing date of birth, facial image and fingerprint, in around two hours. Riscure notes that that the speed of the crack is aided by the Dutch passport numbering scheme being sequential.

When the passport is cracked, you lose all that information, as well as an electronic represenation of your fingerprint. A new driver's license you can get, but new fingers? The centralization of personal information in this way makes the subsequent loss that much more traumatic. After the theft of biometric data, how do you prove that you are who you say you are?

A Deeper Look at the VHA

Via MR I saw this post at Brad DeLong's place reprinting the content from a Krugman piece at the NYT (RR, but included for completeness).

The crux of the point, in case you don't feel like backtracking through all of that, is that Krugman is suggesting that the Veteran's Health Administration is a model of efficient, effective health care, provided by the government to boot.

Last year customer satisfaction with the veterans' health system, as measured by an annual survey conducted by the National Quality Research Center, exceeded that for private health care for the sixth year in a row. This high level of quality (which is also verified by objective measures of performance) was achieved without big budget increases. In fact, the veterans' system has managed to avoid much of the huge cost surge that has plagued the rest of U.S. medicine.

To what can we attribute this, you might ask? Why, to the very fact that it's entirely centralized and run by the government!

The secret of its success is the fact that it's a universal, integrated system. Because it covers all veterans, the system doesn't need to employ legions of administrative staff to check patients' coverage and demand payment from their insurance companies. Because it's integrated, providing all forms of medical care, it has been able to take the lead in electronic record-keeping and other innovations that reduce costs, ensure effective treatment and help prevent medical errors. Moreover, the V.H.A., as Phillip Longman put it in The Washington Monthly, "has nearly a lifetime relationship with its patients." As a result, it "actually has an incentive to invest in prevention and more effective disease management. When it does so, it isn't just saving money for somebody else. It's maximizing its own resources. ... In short, it can do what the rest of the health care sector can't seem to, which is to pursue quality systematically without threatening its own financial viability.

Well, I can't claim to have any experience with theh VHA system. But I'm always a little leary about these "just-so" stories, especially when it involves government bureaucracy. So I just have a couple of points to make about the VA.

First, the potential usage pool is highly restricted. Because they treat only veterans, they don't have to deal at nearly the same level as other hospitals with the surges in demand, the heavy use of emergency rooms as just-in-time care for things that could have been treated sooner and far more cheaply. The drastically smaller size in population isn't a trivial factor; some things don't scale up as well as we'd like, especially under pressure from things like HMOs to prevent expensive unused capacity.

In addition, many of the vets treated by the VA are often covered by other forms of health care. Here are some statistics on coverage by alternative health care regimes for vets. Vets with alternate forms of coverage are clearly opting to alter between them when they see a better deal or better treatment. Non-vets don't have this option. With a unversal payer system, no one would have this option.

On those achievements "without big budget increases": we should account for the fact that the number of vets in the US is both aging decreasing. So as the population treated is going down, the cost of treating them is going up. (Link to 2005 CBO Report: "The Potential Cost of Meeting Demand for Veteran's Health Care.)

More from the CBO report (it's a long excerpt, but eliding too much would be misleading):

VA has had difficulties coping with the large influx of new users seeking pharmaceuticals and outpatient care. Although VA has substantial excess inpatient bed capacity in many facilities, the influx of new enrollees seeking pharmaceuticals and outpatient care has exacerbated waiting times for all veterans wanting to see a VA provider. By the end of 2002, about 300,000 enrolled veterans were on waiting lists for VA medical appointments.

Waiting times have been a long-standing problem for the department. In 1993, the General Accounting Office (GAO, now known as the Government Accountability Office) found that veterans frequently waited eight to nine weeks to obtain appointments at some specialty clinics.(8) In 1996, lawmakers enacted legislation requiring VA to serve veterans in a timely manner.(9) In response, the department initiated a number of actions to address waiting times and waiting lists, including better tracking, better scheduling, and use of a primary care model--that is, coordinated health care delivery through interdisciplinary teams.

Accompanying the rise in the number of veterans seeking care at VA facilities were substantial increases in the annual budget for VHA. Although VA medical budgets were relatively flat in real terms in the mid-1990s, they grew by an inflation-adjusted 4 percent to 10 percent each year from 2000 to 2004. Those budget increases were appropriated by the Congress to fund the rapidly increasing demand for VA health care that followed the change in eligibility rules after 1996.(10)

In part because of the long waiting lists and influx of new patients that VHA could not accommodate in a timely manner, in January 2003 then-Secretary of Veterans Affairs Anthony Principi decided to cut off enrollment of new P8 veterans--those without service-connected disabilities who have income above $25,842 per year (for a single veteran) and above a geographically adjusted means test. Veterans in that priority group who had already enrolled in the system were "grandfathered," however, and could continue to seek care from VA. According to the department, "Until the waiting time for medical appointments can be reduced to an acceptable standard, it would not be in the best interest of those most in need of care for VA to enroll additional priority group 8 veterans."(11)

So, there is some contradiction to Krugman's assertion that this success (of which there does appear to be some) without "big" budget increases. Of course, one Budget Office's "substantial" may not be another man's "big."

More importantly, however, a good deal of the success seems to have come at the cost of restricting access to care. The system was not working well with its current demand, so it was pared down to keep future demand limited. The flexibility to respond to changing demands required shifting care away from specific people. If a single-payer system becomes inefficient and ineffective much like how the commenters on DeLong's site all agree the VHA was a few years ago, would it be able to deny care to subsets of the population while it takes time to retool? Maybe a series of "rolling blackouts'" while the health care network tried to figure out how to expand capacity in time with soaring demand?

I don't dispute that there have been significant, benficial changes at the VHA. What strikes me as less clear, however, is that it has been an unmitigated success of policy. The only cost in restructuring the system hasn't been a "small" budgetary increase for the same or better level of care. And the increasing demand that has put strains on this system before isn't going to be exacerbated much (if at all) by returning vets from Iraq since, as the CBO points out, inpatient care is actually underutilized right now. It's the aging current vet population -- the population whose ailments are among the most expensive to treat in large part because they persist for quite a while -- that is going to test whether or not "one of the best-kept secrets in the American policy debate" is a stunning success or not.

Evolving Institutions

What insight might be gained from considering institutions as an evolving system in the same way we might view an ecosystem? That's the question posed over at the Complexity Blog.

One major focus in political science is the role of institutions, particularly for social choice problems. I was thinking about the relation of individual policy decisions and the institutional framework within which they are made and it occurred to me that the relationship has some analogies to the relationship of species evolution and ecological change. Specifically, analogies exist with regard to the i) time scales, the ii) forces exerted on each other, iii) endogenous stability, and iv) susceptibility to exogenous perturbations.

A thoughtful post, and if you're at all interested in the topics of complexity or agent-based modeling, the site is worth a regular read.

Get Better at eBay

From Knowledge@Wharton, an article on bidding behavior in online auctions:

Would you like to go on an Internet auction site and know how much to bid for a certain item -- and also know that you didn't overpay for that item? How about when you sell an item in an online auction: Would you like to know what price to set that ensures you don't leave money on the online table?

Wharton marketing and statistics professor Eric T. Bradlow can't provide specific answers. But he does offer guidance on the behavior of potential buyers in a new study entitled, "An Integrated Model for Bidding Behavior in Internet Auctions: Whether, Who, When, and How Much," recently published in the Journal of Marketing Research. Bradlow, who is also academic director of the Wharton Small Business Development Center, co-authored the study with Cornell marketing professor Young-Hoon Park. "To the best of our knowledge, this is the first attempt to model formally the behavioral aspects of bidding behavior for the entire sequence of bids in Internet auctions," the authors write.

Here's a a draft of the paper.