Friday, May 4, 2007

Truths, Danged Truths, and Statistics (Mikko Ellila Update)

Followup.

Conservative Beach Girl links to a new post (with portions in English) from Mikko Ellila, in which he states his view for the reasons for the police visit.

This is a bilingual post about Mikko Puumalainen...who told the Central Criminal Police (Keskusrikospoliisi, the Finnish equivalent of the FBI) to interrogate me (the interrogation was later delegated to the municipal police department) for having committed the thoughtcrime of e.g. quoting crime statistics published by Finland’s Ministry of Justice which prove that African immigrants commit more crimes per capita than the Finns do....

[H]e also explicitly said that the government “should intervene especially strongly” in the activities of those citizens who seek to maintain the current anti-immigrationist political climate in Finland by “using publicity”....

Quotes from official crime statistics published by the Ministry of Justice undoubtedly “help maintain an anti-immigrationist political climate” because they prove that e.g. the Somalis commit more than 100 times more (over one hundred times more, as in, over 10,000% more) robberies per capita than the Finns do.

Puumalainen wants to make it illegal to publish crime statistics.

I wonder whether he is going to sue the Ministry of Justice.


After Conservative Beach Girl linked to Ellila's article, she added some thoughts of her own:

I read data from FBI or Department of Justice statistics back in 1996 or so that claimed that if all crime committed by American blacks were removed from the stats, America would have a 5.6% per capita crime rate making us safer than any European nation at the time.

Now, folks, that's not racism and that ain't "hate speech" or "hate thought" - that's getting the data to make a point to support a perspective, whatever that perspective is.


Assuming the statistics are correct, you can (as Conservative Beach Girl notes) "support a perspective" by their use. Ignoring moral constraints for a moment (which will anger my moral atheist readers, but please my amoral atheist readers), it is theoretically possible to support ANY perspective. (For example, take my use of the "Deep Blue Day" song from the Apollo album to prove that Brian Eno is Slim Whitman's illegitimate son.) But the authors of Wikipedia (caveat: I am one of those authors, so you can discount Wikipedia at will) have written the following about "misuse of statistics":

A misuse of statistics occurs when a statistical argument asserts a falsehood. In the period since statistics began to play a significant role in society, they have often been misused. In some cases, the misuse was accidental. In others, it was purposeful and for the gain of the perpetrator. When the statistical reason involved is false or misapplied, this constitutes a statistical fallacy.

Some of these misuses have to do with the data itself (e.g. biased samples), but there are other possible "misuses" cited by Wikipedia:

  • Loaded questions

  • Overgeneralization

  • False causality

Of these, false causality is the most interesting to explore. Let's start here:

Some may bridle at the suggestion [Shane] Doan admits to making — namely, that French referees might be swayed subconsciously by the French-Canadian crowd and the red-white-and-blue jerseys in Montreal. But this, surely, is nothing more than an accusation of mere humanness, not conscious racial prejudice. On Wednesday, The New York Times reported, for example, that a forthcoming statistical analysis of NBA referees (not yet published or peer-reviewed, but checked for mathematical soundness by independent experts) shows that white basketball referees call more penalties on black players than black referees do. Doan, for his part, believes he was perfectly capable of making the same complaints about a game in Calgary officiated by four Westerners.

Note the sentence "But this, surely, is nothing more than an accusation of mere humanness, not conscious racial prejudice." This is one of those "perspectives" that Conservative Beach Girl was talking about. There are several possible perspectives:
  • French-speaking referees are human, not racist, but subconsciously swayed by the familiar.

  • French-speaking referees are racist.

  • French-speaking referees are stupid.

  • Hockey players from teams outside of Quebec are naturally violent, subhuman thugs who deserve to be punished by referees.

And how do you settle which of these perspectives is correct? Via more study. (Of course, statisticians commonly ask for more study so that they can get additional grants. This is, of course, my perspective.)

So why do Somalis commit more crimes than native Finns, or why did blacks in the 1960s have a higher crime rate than the general population? Or, hitting close to home, why do California public schools perform poorly?

Once you've studied this further...what do you do about it?

P.S. If you like my title, here's a Wikipedia article about the origins of the original phrase.

statistics

Sphere: Related Content

2 comments:

Ontario Emperor said...

It would have been better if the original charges against Ellila had been more focused, rather than scattershot (unless they're trying to throw everything against him and see what will stick).

Pity that our self-appointed black leadership in the U.S. isn't hopping on this. But since there's no profit in it for them, they can safely ignore it.

Anonymous said...

Aapo's translations are incorrect insofar as Ellilä does not use the Finnish equivalent of "nigger" ("nekru") in his article, but the equivalent of "Negro" ("neekeri"). Unlike Aapo seems to think, there is a difference between these words in Finnish, just as there is between "nigger" and "Negro" in English. The former is derogatory, the latter merely politically incorrect.