19 comments

[ 2.9 ms ] story [ 56.1 ms ] thread
(comment deleted)
Statistics can be valuable for companies, and they're not very expensive to obtain. For example if you're considering building a Java IDE you want to know how many professional Java programmers there are.
(comment deleted)
in 2007 airport screeners confiscated 1416 guns?!?!

(the 1.1 million knives isn't as perplexing to me 'cuz [in my mind] there's way more ambiguity around what's a "knife" than a gun).

In a reddit Ask Me Anything post a TSA screener mentioned how his area alone found 1-2 guns a month and a few knives/screwdrivers a day.
I had my spyderco folding knife confiscated. I totally forgot it was in my backpack. Apparently that happens alot.
Right, and a lot of people carry small pocket knives on their keychains, etc.
so what is a gun? maybe a lot of those were toy guns (definitely not allowed on a plane).
The lightweight form of "data-driven journalism". Express delta values instead of insight. Use clipart instead of info-graphic visualizations.
My long-running objection to this kind of quicky journalism exactly. No detailed attribution, no drill-down availability, no footnotes. I developed facster.com to overcome these and other limitations, the goal being to provide research-quality stats from the government and other sources, make the information searchable, browseable, and joinable. I haven't had the time to do much work on it since 2005, so the most recent abstract I loaded is from 2003. Also in some migration the search index got damaged. A lot of search terms are missing, but I've boosted repairing the search index on my to do list.
In addition the data isn't normalized. The rate of something could be decreasing year-over-year but due to population increase the quantity continues to increase. Ignoring the normalized data helps journalists create sensationalistic headlines.
As "journalism" I find it lacking, but it might be fun to write a web widget that would put one stat delta per day.
2.4% of people are in jail? Should add that number to the unemployment figures when bragging about how good the economy is (ops, was) compared to France, Germany or the Scandinavian countries...
That figure sounds insanely high. Can anyone confirm it?
It's might be low[1]: in 2007, 1 in 31 adults were in prison[2] - 3% of the population. However, that may include all the people on probation, as wikipedia gives the number in 2009 as only 1 in 136 actually being incarcerated[3], with perhaps five times that being on parole or probation (from one of the graphs, in 2008).

[1] Low by American standards, at least.

[2] http://www.sanfranciscosentinel.com/?p=19911

[3] http://en.wikipedia.org/wiki/Incarceration_in_the_United_Sta...

It didn't say in jail, it said 7.3 millions incarcerated or on probation or parole, which is fairly significantly different.

Most recent statistic I could find (on Wikipedia, I did not drill down to their source) was about 0.75% actually incarcerated. (2.4 million people)

It didn't say in jail

yeah, but being on probation or parole means you're almost in jail, doesn't it?

> yeah, but being on probation or parole means you're almost in jail, doesn't it

No.

Note that you can end up on probation without ever being in jail.