Friday, March 23, 2012

Good Statistics, Bad Statistics

The good:
In other words, when one of Twitter's newest data scientists applies his craft to McDonald's menu, his algorithm automatically extracts the only food on it that any of us should probably even consider eating.
The bad:
An AP "Fact Check" on the correlation of US energy production with gasoline prices takes us to the border of supply and demand before veering off into a comedy club...

Oh, for heaven's sake - the question is, does additional US production result in lower prices than would have otherwise prevailed? If, just to seize an example, producers only ramp up US production in response to shortages and rising prices elsewhere, a simple statistical analysis such as done here will "prove" that more production is always associated with higher prices.
If you have any recommendations for "the ugly" (as they relate to statistics, not to any pictures of me you can find on my blog, MikeAT--and by the way, you're one to talk!), please leave them in the comments.

2 comments:

Mike Thiac said...

As far as the stats bro, you have posted on the budget deficit, cost of Obamacare, California’s budget problems (especially in the pension area) etc that would definitely be classified as the ugly.

As far as pics, beauty is in the eye of the beholder…and you got me beat on keeping the waist line trim I admit. But at least I got my hair. Granted, graying, but it’s still there! ;<)

Steve USMA '85 said...

Good, your example of using stats to support your conclusion

Bad, your example of using stats to support your conclusion but not taking in consideration other explanative data.

Ugly, pretending to use stats to support your conclusion but in reality, you are knowingly using either falsified or mis-represented data. Classic case is the hockey-stick graph of climate change fame.