This page is a random (mostly) assortment of articles I have found that use statistics, along with some ideas on how I may or may not use them in class. Articles are placed with newer to the top.
Comments are open at the bottom, BUT if a comment does not meet one of the two following criteria it will be deleted.
1. The comment must either have an article to share
2. The comment must have some information or idea on how to use an article already listed.
Great article showing bias in hiring for positions in science departments. Full text available in addition to the abstract!
Perhaps there is some bias in the political realm of cellphones and landlines. http://azstarnet.com/news/national/govt-and-politics/article_94d07627-39d6-5ff7-99d0-aab366f9cc98.html
How about using linear regression in the olympics? This is about the long jump specifically, and has a nice video embedded. http://www.wired.com/playbook/2012/08/physics-long-jump-linear-regression/ Here is a place to find the data to do it in class: http://en.wikipedia.org/wiki/Long_jump
Great article on data mining by companies.
How Companies Learn Your Secrets – NYTimes (PDF) and a shorter article that summarizes the NYT article can be found at: How Target figured out a teen was pregnant before her father.
Comparing TV anchors by viewership allows for some interesting comparisons.
You are worth how much?
What the alpha level is for particle physics. It isn’t 0.05 that is for sure!
Timing of birth impacts test scores later in life
How doctors can ethically harness the Placebo Effect.
A statistical analysis of the Grand Slams in tennis and the “random” drawings used to seed the tournaments. It finds some possible easier draws for top seeds.
Does sleep impact grades?
Is wine worth the cost? Can this be done with coke, pepsi, rc cola, etc? Does this fit into experimental design or testing?
A hockey lesson using player statistics. It is written for IB, not AP. It is low level, but might have enough to build up to hypothesis testing at the end of the year, and have enough for good discussions at the beginning. (I like the idea of using a couple of the same projects and build on them as the year goes on.)
Another statistics page on baseball and the fact that there have been 17 no-hitters that are 4 days apart. The most recent in 2011.
There have been a couple of articles on winning lotteries. The first article is about how to win through probability.
While this next article is about how a statistician won by recognizing patterns that were important to the lottery system.
Or how about this one where a stats Ph.D. won the lottery 4 times!
I could see using these as a way to deepen the discussion of probabilities.
I could see using this article for the meaning of “statistically significant” as well as other issues. It is a political article, but not really super biased. Instead it shows how to separate the wheat from the chaff as it were.
http://www.hamilton.edu/news/polls/pundit is the original document mentioned in the article.
I am not sure how to deal with this, but can we do anything with the fact that the study was done, but says nothing?
http://blog.okcupid.com/index.php/the-best-questions-for-first-dates and idea for conditional probability perhaps?
This article from a stats blog about Lance Armstrong’s doping.
there are a lot of links to other articles in there. I can see a couple of lessons on probability and perhaps statistical significance out of it.
On the topic of linear regression, next time after they are done, put this picture up and have them work with it!
How about using z-scores to compare Tiger Woods with Rory McIlroy’s performance at the US Open golf tournament in 2011?
The best first date questions to ask.
Should you consider the color of the car next time you buy one? Does color influence how many accidents you have had? Follow the links and you will get the pdf that has the full report with all the stats.
Here is a possible linear regression experiment. It would take some time to do and create the data, but it is a valid regression line that gives a very high R value.