Leaping Lizards!

 Dealing with an odd-sized year (365.024219 days) is not as easy as you think.  Not only do we have the popular Gregorian Calendar (4-year cycles, with “by-years”), but some geek gurus have proposed the Earthian Calendar (33-year! cycles).  There’s even a connection to Stonehenge.

Update (3 March).  A careful look at the Earthian Calendar site reveals a calendar generator!  This year’s calendar reveals the 2012 leap day as 11 Pisces 0004 Earthian (or Copernican, as the author prefers to call it).

Tips from Gizmodo and The Geek Press.

This WILL appear on a future exam

RTFQ.

Tip from Flowing Data.

A Primer on Clinical Trials

Over on the Scientific American blogs, Dr Judy Stone begins a series of online articles describing the ins and outs of clinical trials.  Most folks have no idea how complicated or time-consuming the process is.

• Molecules to Medicine

Stay tuned, I’ll keep these updated.

Tip from Boing, Boing.

Weekly geekery

My buddy Jaime is always reminding me of cool stuff I see on the web, but forget to pass on.  I’m trying to improve.

Everyday math

• Fibonacci sequences — you do the math.  (Tip from CBS News).
• How algorithms shape our world (Tip from The Endeavour).
• The powers of ten, as applied to the national debt(s).  (Tip from The Parkway Rest Stop).

Health and medicine

• So long, low salt diet (Tip from the Instapundit).

Entertainment

• Conan, rebooted.  (Tip from the Hud, who certifies we’re all adults to Nanny YouTube).

Optimal Grilling

The Rapid Steak Algorithm.  OK, so it’s really an introductory article about operations research, which we at UTSA, in our infinite wisdom, call management science, to keep the engineers from getting their grubby paws on it.

Writer Sanjay Saigal (is that some kind of Houston name?) promises more of this good stuff.

Tip from the Instapundit.

Stupiphany

Developmental ed is the Bermuda Triangle of higher education.

Regression revelation

Here’s a clever reformulation of simple least-squares regression that estimates slope as a weighted average of pairwise slopes.  Too cool!

Heresy in math education

Jan Nordgren pointed me to this dissenting essay.

In fact, if I had to design a mechanism for the express purpose of destroying a child’s natural curiosity and love of pattern-making, I couldn’t possibly do as good a job as is currently being done— I simply wouldn’t have the imagination to come up with the kind of senseless, soulcrushing ideas that constitute contemporary mathematics education.

No wonder I have such a difficult time getting my students interested in statistics.

Anybody got a sense of scale here?

30,000 pigs floating down the river in the recent Aussie floods?  Didn’t that seem like…well…like a LOT of pigs?

Sometimes quantitative literacy is really, really simple, if you just pay attention.

Update (8 February).  Maybe 30,000 isn’t too unreasonable a number!

Not quite ready for prime time

This sounds like a great innovation, using Raman spectroscopy to detect melanoma.  But let’s look a little deeper at the preliminary results.

First, the developers claim that in 274 known cases of melanoma, the device detected all 274 of them, which gives and estimated test sensitivity of 100%, which is patent bullshit; nothing in this life is guaranteed perfect.  A reasonable Bayesian shrinkage estimator might be 275/276 = 0.9964, which is still pretty good.  Second, the developers conveniently omitted any statistics on specificity (how well cases of NO melanoma are correctly identified); I’m sure this was simple oversight*.  Third, and finally, there’s no mention of the prevalence of melanoma in the general population.  Without all three of these numbers, we have no way of evaluating the utility of the device.

But what the heck, let’s do a back of the envelope** calculation.  We have the estimate for sensitivity, and we’ll be optimistic and assume that the other 726 subjects were all melanoma-free and tested negative, so we’ll get a shrinkage estimate of 727/728=0.9986  (which should make it a pretty good test).  Then look up the prevalence of melanoma (also called incidence rate) to get 1/5074 = 0.000197.  Then plug into Bayes’ Theorem to get the posterior predictive value:

PPV = (sensitivity x prevalence) / (sensitivity x prevalence + (1-specificity)x(1-prevalence)) = 0.1251

So if you test positive with this gizmo, you have about a 12.5% chance ( 1 in 8 ) of having melanoma.  I think we need more test data before buying any stock.

Tip from the Instapundit, who should calm down.

* I’m sure there’s an Easter Bunny, too.

** OK, back of the spreadsheet

Update (8 February).  Jan Nordgren has the perfect quote to describe this situation.  I wish I had used it for my title.