A good way of demonstrating the power of illustrating data. And there is coffee. Dunkin' Donuts collected data about coffee drinking habits in the US. See below: Then, designer at en.ilovecoffee.jp turned this hum-drum data into very  p retty data : Property of en.ilovecoffee.jp

This story from NPR is a good example of 1) the media misinterpreting statistics and research findings as well as 2) Type I errors and the fact that 3) peer-reviewed does not mean perfect. Here is a print-version of the story , and here is the radio/audio version... (note: the two links don't take you to the exact same stories...the print version provides greater depth but the radio version eats up class time when you forget to prep enough for class AND it doesn't require any pesky reading on the part of your students).

Triple hilarious with bonus points for being super funny. Damn Avas! Not terribly educational but does illustrate the fact that correlation does not, in fact, equal causation. Property of Bloomberg Businessweek and Vali Chandrasekaran

Darryl Bem on The Colbert Report I love me some Colbert Report. So imagine my delight when he interviewed social psychologist Darryl Bem . Bem is famous for his sex roles inventory as well as his Psi research. Colbert interviewed him about his 2012 Journal of Personality and Social Psychology article, Feeling the Future: Experimental Evidence for Anomalous Retroactive Influences on Cognition and Affect, which demonstrated a better-than-chance ability to predict an outcome. Here, the outcome was guessing which side of a computer screen would contain an erotic image (Yes, Colbert had a field day with this. Yes, please watch the clip in its entirety before sharing it with a classroom of impressionable college students). Big deal? Needless to say, Colbert reveled in poking fun at the "Time Traveling Porn" research. However, the interview is of some educational value because it a)does a good job of describing the research methods used in the study. Additionally, b) h...

While not a statsy blog, H and a H is HILARIOUS ( esp. this blog about dogs and moving . I moved across town and my house broken dog promptly peed all over the house...can't imagine a cross-country move!). However, the entry " Boyfriend doesn't have Ebola. Probably " IS psychometric-y and hilarious. It critiques the FACES pain scale often used in hospitals. NOTE: I absolutely don't own the images below, they belong to Hyperbole and a Half. Property of Hyperbole and a Half The language is NSFW but who gives a fuck about that (see what I did there?). I use this as a discussion board prompt in my online statistics class (which is tailored to professionals seeking their BS in nursing) and those students seem to relate to this posting in terms of their professional life but non-BSN students have been exposed to this scale via trips to the doctor and can discuss its utility, come up with examples when a non-verbal scale is particularly useful, ways to improve ...

Slate's Hollywood Career-O-Matic This story from Slate.com allows students to think about sampling error and data visualization within the context of movies, actors, and directors. This Slate article discusses the movie review meta-webiste  Rotten Tomatoes . This website is of statistical note in and of itself since it compiles the reviews of many, many film critics and then provide at Rotten Tomato score based on this sampling. The Slate article takes this data a step further by providing an interactive  chart that you can use to generate graphs that track a given actor or director's career. Below, the Tomato ratings of the films of Lindsay Lohan. When I use this in class, I ask the students if they believe that Rotten Tomatoes data should be used to set salaries or guide casting. Additionally, I like to provide my students with the 10 Ten Highest Grossing Movies for the year as well as the Top 10 Highest Rated movies (as rated by Rotten Tomatoes) of th...

This is a site I found this summer during the hype surrounding the London Summer Olympics. If you enter your weight and height into the site, it will match you with the Olympian who has the most similar weight and height as to predict your ideal Olympic sport. Needless to say, there are more than two factors that determine one's ideal sport. Which is a great starting point when discussing multiple regression and making predictions. Students can discuss whether or not they've ever played the sport predicted (what is handball? I dunno) as well as list other factors that determine athletic preferences (SES, individualistic vs. collectivist tendencies, body composition, hand eye coordination, etc.).