It's important to see what's going on. I have found ShadowStats (http://www.shadowstats.com) to be a useful baseline because the statistics are transparent (same reason professionals like it: it helps not to have a goal for your numbers). Recommended.

OK, just so I have a baseline for ( Read more... )

"Correlation versus causation in a single graph" | Chris BlattmanChris Blattman


( Read more... )

How else can I describe the latest campaign by Senator Feinstein and a "Breast Cancer Fund" which, oddly, seems to do little other than advocate for the control of things not strongly associated with breast cancer, but which sound bad.

I ought to amplify that last. When discussing breast cancer, the American Cancer Society has this note under ( Read more... )

First, a quick reminder on the vulnerability of science to easy-to-make statistical mistakes: Science News has the scoop. Mainly dealing with medical and laboratory science, it retells a couple of important stories, e.g.

In 2007, for instance, researchers combing the medical literature found numerous studies linking a total of 85 genetic ( Read more... )

Eliot Spitzer (yes, the ex-governor with that problem, now considered rehabilitated by Slate magazine because what he's saying is what they want to hear), has decided that higher taxes don't matter to the economy because of a couple of graphs involving GDP and taxes ( Read more... )

As it turns out, it depends a lot on who you want to count. The key numbers are in table A-12 of the Bureau of Labor Statistics release.

Let's spend a moment with those definitions, using the rationales of the people who created them:

U-1 Persons unemployed 15 weeks or longer, as a percent of the ( Read more... )

As mentioned before, lengthy disclosures degrade, rather than improve decision making.

Today's example: New York City's new law requiring calorie counts on chain restaurant menu boards.

Did it make a difference in what people ordered?

.... only about half the respondents even noticed the calorie counts and only 15% said they influenced their ( Read more... )

There were media reports about the average cost of cell phone minutes, heavily covered this week. Example 1. Example 2.The hard part? The figure: $3.02/minute. Answer these questions for yourself ( Read more... )

In this case, a really pathetic study on drinking and cancer, whose upshot, as the article properly notes, is that you should be cautious about assuming that widely reported studies contain a grain of truth unless your prior intuitions confirm it.

Drink up, ladies.

World to End, Women and Minorities Hardest Hit.

now it's a news story....

Now, with Detroit reeling, many blacks find their economic well-being threatened ( Read more... )