The Wisdom of Crowds Questioned

Slashdot (as an entity) has been whining about the nasty email it gets accusing it of bias.

Of course there is the perception that as in any human filtered media organization, bias happens naturally.

I have been thinking about crowd bias. If you have a crowd of "evangelical christians" for example, what kind of crowd stupidity would come out of it on account of commonly held misconceptions?

Let's take the misunderstanding of the historic origin of "when human life begins" in common usage in the U.S.

In the 1960s, as a part of its general rethinking of liturgy and philosophy, the Catholic Church considered the question of when human life begins. The context was this: Humanity has rights as a group that are different from individual rights. The rights and responsibilities of the group need to be considerate of its members who have no way of making themselves heard. These may be oppressed populations in different locations or they may be the unborn members of the community of humans.

The theologians consulted with biologists to clarify what was then understood as constituting life and came to the conclusion that they couldn't tell when life began. To be on the safe side, they (Roman Catholic policymakers) decided that life begins when the possibility of life exists.

American theologians, taking this as excellent advice, incorporated the new policy into their thoughts.

What we are left with is a conflict between the rights of the family of humans versus the rights of the individual based on a decision so conservative it verges on fantasy.

Gradually, groups who incorporated this into their core belief systems have come to make it a matter of faith.

Catholics, in their admirable love-hate relationship with Rome are somehow able to distance themselves from a good deal of this but evangelicals who are as a matter of belief are in daily direct tow-way communication with God have forgot that it wasn't the result of a divine utterance.

Mandatory Sarah Palin remarks

Sarah Palin Totally Looks Like Tina Fey

On the other hand, she wanted to fire the town librarian for refusing to ban some books she didn't like and built a sports center that cost so much the town COULD have gotten a sewer system. If they wanted it. I guess they didn't want a sewage treatment system. Maybe you don't need one for a town the size of two high schools. Maybe they just freeze it and FedEx it somewhere.

If the Republican ticket wins, Tina Fey will have the best job in town for four years. She can literally spout gibberish and it will be funny since Palin is a Pentacostal.

Accountability in Educaion and the Public Sphere

If you are interested in accountability and bottom-up transformation, here is something we in education should think about.

Most accountability programs are top-down (big stick.) Power is vested way upstream in a licensing authority of sorts such as a state's department of education rules that essentially decertify a professional under a specified set of circumstances that are documented under a routine review and filtering process.

There is a huge volume of writing and public assertion about giving the public "choice" but that is usually defined in an on/off paradigm of either using a public institution or leaving and going to a private institution. Choice means giving the public the choice pundits want to give them. I could rant about religious fundamentalism here and the damage it has done to the corporate mentality of the American public. But I won't do that.

Instead, how about empowering the public with tools for change?

This is a post at Howard Rheingold's site Smart Mobs titled:
Citizens use YouTube to keep gov’t in check

You have seen it working with the Los Angeles Police. But as people realize the benefits of social network software giving a low threshold for attention, it is possible to extend to other public spheres. Public nuisance is certainly a good example, but opening the classroom like a fruit is another interesting example.

We already have open curriculum to the extent we use Moodle where authentication can be given to anyone as opposed to a centrally managed system like WebCT. We have toyed with putting webcams in classrooms.

To tell the truth, I believe that student posting of school events whether in class or not is inevitable. All I am really talking about is giving it a forum.

Let's say Rate My Professor decided to allocate a hunk of bandwidth to video and students could capture, edit and post within minutes in a discussion format.

Whoever gives shape to the first marketplace entry will frame the discussion.

Why I didn't buy a new iPhone how Android gets my vote

I'll admit it. I didn't buy a new one. The old one was nice but when it spent a half hour in the washing machine and I went back to my first generation iPod (waiting for 7-11-08) I found I could still listen to the podcasts to which I am addicted. I wasn't *unhappy* enough to go back.

THEN I found out the data charge would be thirty dollars a month! Frankly I am not willing to pay that much to look at a four inch screen. A dollar a day isn't much but I only use the data service once a week unless I'm out of town. So that's seven dollars per use.

I will probably reevaluate in 2009 and see how things are going. In the meantime Android will launch on new phones for multiple service providers.

I was going to use my iPhone in my Interactive Design class as a guinea pig/sacrificial phone. But after consideration I realized Android is more practical for my students since they will be able to afford their own Android capable phones next year. This will vastly improve the social aspect of the class.

The social interaction that surrounds handheld devices is worth of spending time studying.

Global Search and Stress

This post:"Is Google Making Us Stupid?" by Craig Newmark proposes some interesting observations about the interaction of wealth and personal well-being. I would like to make some observations on search technology and the way we interact with it. Newmark is quite right in the way he points out the deleterious effects of our interaction with technology. The original article with the same title ran in the Atlantic.

There are quite a few subjects in the post that deserve treatment. But one theme that runs through is choice.

If you are conscious of your habits, it is not difficult to immerse yourself in a deep and lengthy work. However, people and students in particular choose not to read information in full context. How much of it is the way they have been taught to read and how much is the style of writing is debatable.

To tell you the truth, the last sentence of the post is probably my favorite:

"(I note a third hypothesis, also related to economics. Since information has become more important and more valuable, some authors have gotten much better at producing it. High quality information lowers patience with the lower quality forms. Once you've read a really well-written microeconomics textbook, say, it's much harder to wade through a poorly-written one, or a poorly written text on any subject.)"

Television is a good parallel that I can use as an example of why I think this is true. OMG, I didn't mean that. I am giving an example of production volume, not quality. Don't worry, I will get to finding quality a bit further on.

As with the Internet, cable and satellite bandwidth far outstrips the content a given individual wishes to consume.

Viewing habits are a good indicator. I survey my freshmen every semester on how many network channels they watch in any given week.

In a class of 25, the median is from 7 to 9 with very small variation. There will be one student with no television, one near 15 and one over 20.

This seems at odds with the availability of several hundred alternatives. Because I am a technology and society guy, I only survey news content viewing as a measure of choice variation. Again, there is very little spread. 80% of the average class watches one of only three channels out of approximately 25.

How much of the choice is due to poor interface? A Finnish researcher claimed in 2005 that our interactive (haptic) interface hasn't changed in over 100 years. This was based on his comparing current technology with mechanical arcade games that use pressure sensors to deliver feedback from the turn of the previous century.

In the case of television, development capital is scarce because of the low chance of economic success. Consequently both content and quality are scarce in comparison to bandwidth. (Think repeat of last season yet again.) In the case of Internet content, the development capital threshold is so low my students can afford it. Content creation has been democratized to an amazing extent. This makes the process of making a reading or viewing choice so difficult, we need the help of a search engine. Quality has to be treated differently as you will see.

The famous failure of full text search to live up to expectations is also to blame for a good portion of the reading statistics. Searches turn up hundreds of thousands of hits, a good number of which must be called up in order to eliminate them from consideration. This is the so-called "bounce rate" but it is not indicative when a reader finds exactly the page he is looking for. Bounce rate is really a measure of depth of site penetration, not content quality or relevance.

So Google does have something to do with it (!) because Google is unable to search for a concept that can be described in many ways. Take for instance a full text search for an emergent concept that people label or describe differently. But what the heck, it isn't magic, it's just a computer. The user had better be able to bring something to the table.

To some extent, that's why Google bought GapMinder. You can see the nations of the world compared by average income, infant mortality, longevity, and GDP over time. It is easy to see that if a population is healthy, average income will increase more rapidly than it does in a country where health care remains poor while economic development is happening.

GapMinder is an example of the semantic web and how it may well cause an accelerated discovery revolution comparable to that of the rapid discovery revolution of the 1600's.

Filtering noise in an effective way will be the next challenge that allows stress to subside. But the challenge is huge. How do you decide whether to exclude news from Fox or the Competitive Enterprise Institute in a meaningful way?

I feel that there are two things that will reduce stress when it comes to overload. First, open access to professional communities. Second, semantic utilities with tiered relationships. (Tiered in terms of technical knowledge and links outward to either collections of data or critical evaluations of relationships.

Is this just another whiz-bang computer nerd technology solution? That's what I heard in my SIG meeting when I proposed we sponsor a DOE symposium on one of their collaboration pieces. But Raymond Orbach (DOE) and Abe Leiderman (Deep Web Technologies) have produced good work with federated data previously and I see no reason to think that extending it a bit is unreasonable.

Hyperlinks and search results are seductive in the way wandering in the library is. Debating the percentage of the population susceptible to wandering off topic in a given circumstance is probably not productive. Again, it's choice.

Does Google make us stupid? To the extent that using a tool without knowing how full text search and popularity ranking works can produce ineffective and flawed work. (Student papers. rofl rofl) Using other search techniques such as reading full context and gauging professional reputation can help stabilize the work. In other words, making good choices is central to success in search.

NCLB (FCAT) Broward County Schools

NCLB testing (aka FCAT) results in Google spreadsheet form. Broward County schools ordered by percentage of students qualifying for free and reduced lunch.

The "poverty percentage" column is ordered from lowest to highest to highlight the test results in low poverty communities. All schools have been included regardless of student age. Without working any further, it appears to me that minority percentage may be a stronger indicator but if you're high in both columns, an A or B looks really hard to do.

I might order it differently from time to time so don't be surprised. It should not be hard to tell.

Negative Tagging for research

One could propose a tag or set of tags for publication in a dB (say PubMed) for "negative results" that might be a nice way to search so you could find them more easily.

One might also encourage publication of negative results since the barrier to publication is so low on the Internet. (Peer review not omitted of course.)

Take THAT Ben Stein!

News from the BBC:

...the Vatican is organising a conference next year to mark the 200th anniversary of the birth of the author of the Origin of Species, Charles Darwin.

Father Gabriel Funes, Vatican Astronomer made the announcement and included speculation on the spiritual state of possible alien races.

This is a good example of how theology should advance. It opens speculation on the nature of spiritual subject matter and does not glorify text written by persons who lived hundreds of years before the nature of space, time, and matter were explored in a meaningful way.

Laptops and Students - To Be OR Not To Be -

... That is the question. Whether 'tis nobler in the mind to suffer the slings and arrows of outrageous fortune...

Gentle Reader, kindly peruse the previous two postings before tackling this one. They will give you some background knowledge that should put you in a frame of mind to accept this particular bit of opinion. It *is* opinion of course since it is not backed up by research.

Probably the most important thing to recognize when comparing the XO or One Lapotop Per Child (OLPC) and the U S district based projects is the nature of the combination of network and resources.

Not just the hardware. The rest is a lot more important.

The whole thing is a huge tool and people, especially kids, will figure out how to use it. Yes, they will hack it.

If you ask Nicholas Negroponte if he thinks the kids in Peru will hack any of his systems, he will say "I sure hope so!" It will be disappointing if they don't.

1. This huge tool must be transparent to use. This means that the tool becomes irrelevant, like a pencil or a pen. The fact is that it is largely transparent to students, but not to everyone else.

2. The tool exists as a cloud of resources to be used at will. It is not time or location dependent. So when it doesn't fit into a particular frame of reference, just forget about trying to make it fit. The tool, by its nature, works.

If changing the location of an activity is important, consider that everyone involved in using the tool can and will leave the location at will.

If changing the tool itself is important, consider that everyone involved can and will repurpose the tool literally in a few seconds. (If you can supervise children with scissors, you can supervise a class of technology users. That may sound challenging, but art teachers do it every day with very few fatalities.)

Finally, think of the scope of the community involved and the connections available.

In the U S, proxy filters are common. In Finland, Sweden, and Denmark I hear the children are expected to behave themselves without nanny software. Granted, Finland has very high test scores in Math and Science.

In order to control and guide students who have this kind of power, the environment has to be carefully curtailed in certain ways without stifling creativity and access in an unreasonable way.

Consider the typical situation envisioned by the OLPC program. It is a compact community having few members with no preexisting Internet access. The school provides a satellite uplink and lesson resources. The student's laptops create a mesh network that presents the same face to the users no matter whether they are inside the school building or outside.

This means that instead of being cut off from the school environment when the day is done, the school follows the child home and provides a community with teacher support and peer support.

Let me mention one of the criticisms that OLPC communities have. Spouses of teachers complain that their time at home is absorbed by students. After letting that sink in, consider that one of the hallmarks of a high quality education is long periods spent with professors.

Let me close this by quoting from Seymour Papert's article about Bode Miller

Consider a simple incident: a 3-year-old goes to a shelf, pulls off a cassette and loads it into the VCR. People are wrong to be amazed at the technological capacity of the child. Getting dressed and playing with many traditional toys are actually more complicated than working a VCR or even clicking a computer program into life. What is remarkable is that the child is able to make a decision to spend the next half hour immersed in a topic of choice...

That children are learning to find independent ways into knowledge is wonderful and necessary. But it poses a challenge for parents and indeed for the way we think about school. In the past school had to provide knowledge. In the future schooling -- and parenting -- will have to be about developing the ability and the judgment necessary to use knowledge wisely and critically.

I hope the link with what I said earlier is obvious. Students will figure out how to use technology better than their elders.

The OLCP project is designed to support precisely the parts of networked community that are useful to discovery and education. This means the XO laptop is designed precisely to be a particular subtype of tool for a reason that the majority of technology projects in the US are evidently blind to it.

Una Laptop por Niño - One Laptop Per Child

The philanthropic effort dubbed the $100 Laptop has not met its grand initial goals. But its first deployment, in Peru, may turn skeptics into believers.

Here are a few quotes pulled for the article.

The laptops are headed to 9,000 tiny schools in remote regions such as Huancavelica, in the Andes, an arduous 12-hour bus ride over rocky roads southeast of Lima, and villages such as Tutumberos, in the Amazon region, days away.

The computers come loaded with 115 books--literature such as Mi Vaquita, about a rare porpoise, but also classics, like some of Aesop's fables, novels (at least one by the Peruvian writer Mario Vargas Llosa), and poetry (including verse by the early-20th- century Peruvian poet César Vallejo). The laptops' flash drives also store introductions for teachers, reading-comprehension programs and other educational software, a word processor, art and music programs, and games, including chess, Sudoku, and Tetris. The rugged, low-power hardware includes a camera that can capture video or still images. The computers are Internet ready and can wirelessly relay data to one another.