Carronade The Yankee Sailor Carronade

The Sea is a choosy mistress. She takes the men that come to her and weighs them and measures them. The ones she adores, she keeps; the ones she hates, she destroys. The rest she casts back to land. I count myself among the adored, for I am Her willing Captive.

I've relocated to a new Yankee Sailor.

Monday, April 03, 2006

I Thought This Horse Was Dead, Part III

For background on this extended discussion, see parts I and II.

OYE responded to part II, and the relevant parts begin with this:

Given the constant evolution of and rapid turnover in our military, Carter Administration enlistment data are completely irrelevant today. [They've done their twenty already.]
It's correct that most if not all of those servicemen are no longer serving, but citing the figures provides something OYE desparately wants to conceal: context.

The 1991-2003 enlistees (of whom less than 1% were Category IV - click here) have done great things in Afghanistan, Iraq and elsewhere. Would you want a high percentage of Category IVs in your unit?
Twice in my career I've had the pleasure of serving as a First Lieutenant aboard warships, and that means one thing: Boatswain's Mates. Out of the 80-125 Sailors in my departments, usually about 15-25% were Cat-IV enlistees, and less than 5% of them gave me any problems. Would I have trusted them with trigonometry? No. But they were almost universally proud, hard working and exceedingly capable technicians and mariners. Would I do it again? In a New York Minute.

We agree that the October figure [of 12% Cat-IV's] may or may not be relevant by itself, precisely because we don't expect the October cohort to be a representative sample of the entire year's recruits. We'd accept your characterization of the October figure as an "outlier" if the Pentagon would release the same information for November, January, February, and now March. Statisticians would not call it an "outlier" without that information; neither should you.
Given that the cap on Cat-IV's for 2005 was not more than 4%, it is statistically impossible for that figure to be representative of the entire year. It may be an ominous predictor for 2006, but, assuming the military did not exceed its cap and fail to report it, for the calendar year it was an outlier.

I certainly expect higher quality reasoning supporting the decisions you make on the job to defend our nation.
Higher quality reasoning? Maybe that means they can't understand data and arguments without any "spin."