A very difficult topic with computers are significant digits. This is summarized in the comic below:

Saturday Morning Breakfast Cereal via kwout

Saturday Morning Breakfast Cereal

Found via Flowing Data

This article from the New York Times assumes that household consumption is a better indicator of Americans’ living standard Than income. This idea is based on the Bureau of Labor Statistics that say that the bottom fifth earned just  $9,974 but spent an average of $18,153 a year. I copied below the explanation provided:

How is that possible? A look at the far right-hand column of the consumption chart, labeled “financial flows,” shows why: those lower-income families have access to various sources of spending money that doesn’t fall under taxable income. These sources include portions of sales of property like homes and cars and securities that are not subject to capital gains taxes, insurance policies redeemed, or the drawing down of bank accounts. While some of these families are mired in poverty, many (the exact proportion is unclear) are headed by retirees and those temporarily between jobs, and thus their low income total doesn’t accurately reflect their long-term financial status.

Taking consumption as indicator instead of incomes, reduces the gap between top fifth and  bottom fifth to 4 to 1. This ratio is 15 to 1 if incomes is used as indicator.

You Are What You Spend – New York Times via kwout

Alec Saunders published an interesting post that describes how to do some quick data mining with Facebook. The starting point is to click on the advertisers link at the bottom of any Facebook page. Follow the steps as described by Alec up to the audience choice. That’s where the data mining can be done. For instance  Facebook reports that there are 1,053,740 people located in France:

If more accuracy is required, we may select only women. There are  319,600  Women located in France:

We can do the same for Men. There are 321,440 men located in France.

There are slightly more men, but there is almost a perfect balance between men and women. The strange thing is if we add Men + Women located in France, we get 641,040 (and not 1,053,740). This means that about 400,000 people didn’t complete the sex information in their profile. This point shows the limit of accuracy of Facebook data mining.

Another interesting finding is the difference between single and married people. There are 166,280 Singles.

And there are 56,540 married people.

We can explain the fact that there are more single people because there are more young people (under 25 so not married yet). It will be interesting to check those figures in a few months to see how the network evolves.

According to this article from eMarketer, Facebook is now the fourth web site worldwide in terms of traffic. The graph below is taken from the article:

Found via FaceReview.

Hans Rosling is a genius of statistics. He knows how to bring out insights from boring statistics. The following video show one of his presentation at TED:

In a follow-up to his now-legendary TED2006 presentation, Hans Rosling demonstrates how developing countries are pulling themselves out of poverty. He shows us the next generation of his Trendalyzer software — which analyzes and displays data in amazingly accessible ways, allowing people to see patterns previously hidden behind mountains of stats. (Ten days later, he announced a deal with Google to acquire the software.) He also demos Dollar Street, a program that lets you peer in the windows of typical families worldwide living at different income levels. Be sure to watch straight through to the (literally) jaw-dropping finale.