Lingo (and more)

The Minitab blog
True statistics geeks may want to check out the "Minitab Blog"
on the software firm's Web site. Bloggers write about serious
matters like assessing the Bureau of Labor Statistics measures
of employment growth, but will also post on topics like "Using
Data to Make Wedding Planning Easier," "Do NFL Teams Have a
Greater Home Field Advantage on Thursday Night?," and "Want a
Raise? Move to a Smaller, Colder State." (The blogger found that
"there are statistically significant relationships between median
household income and both population per square mile and
average winter temperature.")

The rise of "analytics"
Increasing numbers of master's programs in "analytics,"
many as part of business schools, testify to the growing
importance of data-analysis expertise. Programs
often combine statistics with studies in such fields as
information systems, operations research, computer
science, management, and marketing. Northwestern,
Carnegie Mellon, the University of Texas, and North
Carolina State University are among the schools
offering full-time master's programs.

What's in a departmental name?
Kenyon's Math Department recently changed
its name to the Department of Mathematics
and Statistics. The change recognizes the
growing importance of data analysis in many
fields, emphasizes the department's strong
offerings in statistics, and tells prospective
students that they can indeed study statistics
at the College.

Hoist a pint to the "t-distribution"
One of the major tools in data analysis, the t-distribution (a cousin
of the bell curve, used with small sample sizes), was discovered
by a chemist for the Guinness brewery. William Sealy Gosset
(1876-1937), who had studied both mathematics and chemistry
at Oxford, was seeking more accurate calculations involving raw
materials like barley and yeast. When he published his findings in
1908, he skirted a Guinness ban on employee publication by using
the pseudonym "Student." So his discovery is often known as
"Student's T-Distribution."

Lingo: confounding variable, correlation/causation
A classic example: Data analysis has shown a strong
correlation between ice cream sales and murder rates. So,
do murders cause people to buy ice cream? Does buying ice
cream cause people to become killers or murder victims?
An important lesson in statistics is that correlation
doesn't necessarily mean causation. In this case, there's a
confounding variable: the weather (or temperature), which
correlates with both of the other variables-there are more
murders in the summer, when ice cream sales are also high.

Stats stuff
Snipes has a nerdy selfrefererential t-shirt as well as a collection
of stuffed "distribution plushies" like the standard normal
distribution below. (Her students named it "Norm.")