Math 216: Statistics for Engineering

The researcher interviewed here is familiar to me--he's done some interesting things with cryptography to secure information about victims of human rights abuses. Here, he talks about the statistical challenges in estimating death tolls. Very interesting stuff.

I know the interviewee here from my cryptography class. He uses advanced encryption to keep secure information on witnesses of human rights abuses. In this interview, he talks about the challenge of estimating how many people are killed in various conflicts. He keeps the stats pretty light in this interview, but I think you can still get an idea of how statistics might help with these kinds of estimations.

Axiomatic

A variety of actual graphs from media used in a course

View complete Kentucky Wildcats basketball stat leaders from 2011-12, including points, rebounds, assists, steals and blocks.

Scientific evidence for popular dietary supplements.  Turns out there isn't a shred of evidence to support that Vitamin A is beneficial.  Based on human, randomized placebo-controlled trials. 

Here's a (new to me) blog all about data visualization.

Michelson's Light Experiments

;)

Unfortunately, not good dataviz on this site, but some of the linked documents do have some charts and scatterplots

This is a database from the NIST webbook

kentucky's (my state's) manufacturing statistics. 

The "U.S. National Bank Notes, 1864-1935" database was assembled by Franklin Noll as part of a study on National Bank note production by the Historical Resource Center of the Bureau of Engraving and Printing. As such, much of the data relates to production; though issuance, redemption, and circulation figures have also been gathered in an effort to help explain variations in production.

This would probably be an axiomatic notion of probability, as it places the approval of past presidents with the current president. If it were taken from the perspective of whether or not President Obama will be elected for a second term, though, it could be seen as subjective probability statistics. 

data does not load correctly with Internet Explorer, please use any other browsers to view

datasetof statistics on what appears to be every college basketball team from years 2003-2012

and here's one guys analysis of the 2010 statistics and some of his predictions, just about all of his predictions ended up wrong except that he predicted Vanderbilt was the most likely team to be upset in the first round....there's a lot of interesting sports stuff on the rest of the harvard sports analysis site worth checking out too http://harvardsportsanalysis.wordpress.com/2010/03/17/predicting-ncaa-tournament-upsets-favorite-vulnerability/

empirical probability (skip to 1:00)

Note the lack of pie charts here...

anyqs How was the chance of winning calculated based on the percent voting range? I feel like theres an easy formula for this, it's just not immediately obvious (to me anyway).

anyqs This is a very interesting infographic because I use both diig and reddit and most of the information follows a pattern I have noticed. But I feel it could be better if it also shows the number of users that use both sites.