Social Science Statistics Blog

Just simple spreadsheets and charts, but I've pulled together some interesting government statistics at: http://www.supportingevidence.com/Government/government_landing.html.

Check out the pareto chart on U.S. Government Agency budgets. It really makes our investment priorities crystal clear.

Scott Gibson
www.supportingevidence.com
'worth a thousand words'

I think your reference to "relatively uncertified" intermediaries may be missing several points. First, there are many intermediaries of public data in the mainstream media. Are they certified or uncertified, and by what standard? Second, there are many purveyors of public data working in government agencies. Are they automatically to be considered "certified" if they are reporting agency-generated data (even if it often highly selected or even propagandistic)? Third, there are many intermediaries in the commercial world (financial analysts, advertising and sales professionals) who draw on public data to interpret market conditions or make a sales pitch. On what basis can one classify them as certified or uncertified? Fourth, there are many intermediaries in the area of sports statistics -- for example, sabermetricians in baseball such as Bill James and Tom Tango. These are among the most credible people in the statistical analysis of baseball. Virtually none is "certified" in the field but many have sterling reputations among their peers.

Long story short, I don't know what "certified" means or that it is even a useful term. What regulates the quality of the intermediation of public data is critical reading and "the market of ideas" (competitive analysis and reinterpretation of accessible public data). And that doesn't only happen in professional journals with a peer review process. It happens in the blogosphere as well, as analysts develop credibility among their readers (or competing bloggers) for their insightful or persuasive analysis of public data.

Scott, your site looks interesting. Have you thought about sharing your work on Many Eyes, either instead of your site or in addition to it? You can post your data, create visualizations, and write commentary on it -- and then get comments from others, even their visualizations of your data. You could then embed your visualizations on SupportingEvidence if you'd like. It just seems better to take advantage of and try to build an existing community (with great tools!) than to be out on your own like that.

Will, you're right, I wasn't clear what I meant by "certified." I think basically I meant "credentialed," in the sense that some institution (university department, think tank, consulting company, media outlet) has bestowed the seal of approval. I definitely did not meant to say that people without those credentials are not credible. My point is that when the cost of accessing data and publishing analysis about it goes down, more of the intermediaries will be people without these credentials. You're right that these new intermediaries develop reputations and ultimately new credentials (e.g. Poblano, blogger for DailyKos) so it's not quite right to say that the new intermediaries are "uncredentialed." I just meant that intermediaries would be coming from new places. And I totally agree with you that the diminishing importance of the old credentials (eg academic journals) is a good thing -- it's progress when the market of ideas is working.