Monday, October 27, 2014

Nell Greenfieldboyce's "Big Data peeks at your medical records to find drug problems"

NPR's Nell Greenfieldboyce (I know, I thought it would be hyphenated as well) reports on Mini-Sentinel, an effort by the government to detect adverse side effects associated with prescription drugs as quickly as possible. Specifically, instead of waiting for doctors to voluntarily report adverse effects, they are mining data from insurance companies in order to detect side effects and illnesses being experienced by people on prescription drugs.

Topics covered by this story that may apply to your teaching:

1) Big data
2) Big data solving health problems
3) Data and privacy issues
4) Conflict of interest
5) An example of the federal government pouring lots of money into statistics to make the world a little safer
6) An example of a data and statistics being used in not-explicitly-statsy-data fields and occupations

Free American Psychological Association style tutorials/quiz

Here are two free, Flash tutorials about APA style directly from APA. The first tutorial is provides an introduction to APA style, while the second provides a list of changes in the 6th edition.

And here is a free quiz on reference alphabetization, also from the APA Style Blog (you can also download the quiz in PDF format for in-class use).

Also, don't forget on these resources (1, 2) for help crafting results sections in APA style.

Monday, October 20, 2014

Quoctrung Bui's "Who's in the office? The American workday in one graph"

Credit: Quoctrung Bui/NPR
Bui, reporting for NPR, shares an interactive graphs that demonstrates when people in different career fields are at the office. Via drop down menus, you can compare the standard work days of a variety of different fields (here, "Food Preparation and Serving" versus "All Jobs").

If you scoff at pretty visualizations and want to sink your teeth into the data yourself, may I suggest the original government report entitled, "American Time Use Survey" or a related publication by Kawaguci, Lee, & Hamermesh, 2013.

Demonstrates: Biomodal data, data distribution, variability, work-life balance, different work shifts.

Tuesday, October 7, 2014

Free webinar on Simpson's Paradox teaching example/Bayesian logic for undergraduate statistics

Attend CAUSE Web's free Journal of Statistics Education webinar on 10/21/14 to learn about 1) a classroom example  of Simpson's Paradox as well as 2) ways to incorporate Bayesian logic into undergraduate statistics courses.

More information on past JSE webinars available here.

Monday, October 6, 2014

Mara Liasson's "The challenges behind accurate opinion polls"

This radio story by Mara Liasson (reporting for NPR) discusses the surprising primary loss of former Republican House Majority Leader Eric Cantor. It was surprising because internal polling conducted by Cantor's team gave him an easy win, but he lost out to a Tea Part favorite, David Brat. The story goes on to describe why it is becoming increasingly difficult to conduct accurate voter polling via telephone and the internet.

Some specific points from this story that teach students about sampling techniques:

1) Sample versus population: One limitation of polling data is the fact that many telephone call-based sampling techniques include land lines and ignore the growing population of people who only have cell phones.
2) Response rates for political polling are on a decline, making the validity of the available sample shrink.
3) Robocalls, while less expensive, have no way of validating that an actual registered voter is responding to the questions. Additionally, restrictions on placing robocalls to cell phones (but not land lines) create more difficulties for pollsters.

I used this example as a way of introducing sampling error and, eventually, distribution of the sampling mean/standard error.

I emphasized that Cantor's polling had been conducted by high-end polling firm McLaughlin and Associates (source). I also explained just how powerful Eric Cantor was and if an extraordinarily rich and powerful person using a very expensive polling firm has to contend with sampling error, then what does that say for the rest of us?