Stats Made Easy

Practical Tools for Effective Experimentation

Sunday, June 25, 2006

Statistics show slump after workday lunch, but siesta not suggested

Ty-phoo tea commissioned a study that shows 2:16 pm to be the time when UK workers hit their low ebb. One concludes that this would be just the time for a cup of tea or other caffeinated beverage. The results, based on a sample of 2000 people, indicated that energies peaked at 9:23 am and 4:31 pm (time to go home!). I wondered if warmer-climed countries that take mid-afternoon siestas may be on the right track, so I did a quick search for studies on this. However, the one study I found* reports a significant increase in the risk for heart attacks: "Compared to controls, [victims] were more likely to take daily siestas (44% versus 35%, P = 0.01), and spend more time per siesta (1:07 ± 0:04 versus 0:54 ± 0:04 h:min, P = 0.002)". Make mine a coffee please -- no cream or sugar, only black -- thicker the better. I see in this Scottish Heart Health Study that coffee is better than tea for those (like me) who suffer from coronary disease.

*Siesta and the risk of coronary heart disease: results from a population-based, case-control study in Costa Rica

Tuesday, June 20, 2006

Dewar's master blenders make monkey out of clueless mixer

I've just capped off two weeks in Great Britain by touring Dewar's
World of Whiskey
at their Aberfeldy Distillery in Scotland. They
let me taste their famed White Label blended brand, which
according to Dewar's "Never Varies." I must say that I could not
tell the difference of it versus their Aberfeldy Single Highland
Malt or the 12 year old Special Reserve -- they all burned me
through and through (I do not normally drink hard liquor). This
lack of subject matter knowledge may explain why I flunked Dewar's
computerized blending challenge and prompted the software to say
that my whiskey would be good for "stripping paint" and that "a
monkey could do better and would probably be easier to train!" If
only I had brought my computer along to use Design-Expert
software's marvelous statistical mixture design features. It
would have helped me match White Label with the correct blend of
the six components: Highland earthy malt, Island salty malt, Islay
peaty malt, Lowland malt, Speyside sweet malt and the spirit base
-- raw grain whiskey. I wonder how long it would take a monkey to
uncap a beer -- I think I could top it at this task. Anyways,
after this debacle trying my hand at blending whiskey, I may have
a brew or two.