Friday, December 08, 2006

Swivel already somewhat useful

Curse me for being an eternal pessimist. I've not yet managed to shake the perpetual cynicism and supercilious scornfulness that is the academia I've recently escaped from. Or something.

Anyway, I shouldn't have been so sour on Swivel's preliminary test site. Investing a little more time in it, it's apparent that it is already useful--at this point to the extent that you find a title that appears to have what you're looking for, and then in the bottom left of that title's page, you click on "View on page" in the "Source data" box. The hyperlinked headers at the top under "Data", I maintain, are clunky, confusing, and pretty useless.

I experimented by uploading data I'd been using to look at the relationship between the net worth of Congressional representatives and how they vote on immigration issues. As a group, the five wealthiest Senators receive a D+, while the five most modest Senators earn a B. For GOP Senators, wealthier representatives are more supportive of open-borders (a moderate but statistically significant inverse correlation of -.32), probably due in part to having been around longer and consequently becoming more beholden to big money special interests and also to having acquired their own wealth in the business world (or coming from families who have), therefore appreciating the (short-term) benefits to the bottom line that slave labor can provide.

It does allow users to upload data from a few sources (including Excel in the limited CSV format), and it similarly allows users to download them in the same formats. By tagging the uploaded data in an accurate way, users are essentially going to have free access to an enormous database containing all kinds of stuff. Swivel is going to have to figure out a way to make it easy to sift through, so that when I'm searching for income per capita I'm able to narrow it down to fewer than 2,400,100 matches.

I can't figure out how to create my own graphs using my data, but Swivel created a few automatically. Two of them are useless (Swivel apparently crosses columns randomly), but one of them is somewhat informative.

Steve mentions that intellectual property rights aren't going to be preserved when handed over to Swivel. Seeing the litigious mess that Youtube, Napster, and other filing sharing services have run into, the site is prudent to try and get the message that it's going to operate as 'open source' out there as far ahead and as emphatically as possible.

Still, if your livelihood is based on this sort of thing, any data you release are going to find their way onto the site. As a curious amateur, I echo the words of Harry Truman:

It is amazing what you can accomplish if you do not care who gets the credit.
Easy for me to say!


Brian Mulloy said...

Thanks for the follow up post. We still have a long way to go. Please stay critical.

About the IP issues. We've had a ton of feedback about this. And frankly our terms of use don't match or philosophy. We're working on that now and should have good news to report very soon.

Ron said...

you should drop the good chart from Swivel into your blog so we can see what you're talking about.

if you click on the blog button under your graph, you get a really nice "embed this graph" link to drop into your blog.

I'd love to see the data and the argument side by side.

crush41 said...


I feel so guilty as a 'critic' (an enthusiastic one mind you!).

You're working on something incredible, I'm whining about it.

I'll be eagerly awaiting the good news.

crush41 said...


I'm able to see the graph when I click the link from a unique IP. I wonder why it won't show up for you.

Anyway, I've included the graph in the post.