Monday, August 01, 2011

The Bell Curve circa 2011

In the scientific community there are primarily two types of research: Directed and non-directed. Directed research typically involves noticing a problem and working towards solving it, such as curing cancer or breeding a seedless watermelon (guess which one we've cracked?) Non-directed is simply research for research's sake, such as the last time you were fascinated by the contents of your own bellybutton [The nature of navel fluff. (Medical Hypotheses) This is real, BTW]. It is rare, however, to see the second form in conventional newspapers - often the jargon is impenetrable and drawing independent conclusions can be frustrating early in the morning. [Nonacyclo-docosane, a Bastard Tetramantane (Journal of the American Chemical Society)also real]

The Edmonton Journal, however, is boldly going where no one has cared to go before with a map of ethnicity in our fair city, just in case violent racists were having trouble deciding which streets to riot. "Got a beef with purple people? 50th and main!" (In a sad coincidence the map icons even resemble targets) Through census information from 2006 which asked people to identify their ethnicity, they formed maps of Edmonton, displaying where people settled. In case that dubious concept wasn't upsetting enough, they tried to expand on the premise with some uncomfortable conclusions positioned as interesting factual tidbits. "The maps show many Africans concentrated around 107 Avenue in Queen Mary Park, as well as Beverly Heights, Sifton Park near Clareview and in the northeast, Kilkenny and Evansdale, all neighbourhoods with lower than average median household incomes." They just put that there. 'What do you think of that?' it asks, one eyebrow raised and a supercilious smile worming its way onto the stage. "Lower than average median household incomes" it says, staring so hard its eyeballs water.

I can understand some hype for their little map, but it's just a point of interest not the Rosetta stone for housing: " Those patterns reflect the length of time since the first immigrants arrived, how they’ve adapted to the city, and what values they prioritize." I am unsure how the fact that people populated a certain region indicates much about their values, but I have a cup of tea leaves I would love the Journal to look at. In a crazy twist it looks like these neighbourhoods  "shape large areas of the city." Freakish! Other large areas of the city are made of fluff and abstract concepts. I can not fathom how no one noticed that perhaps the people that perfer to live closely to others that match their ethnicity are also more likely to self-identify with that ethnicity on a survey.

The last failing in the map is that it confuses correlation with causation - a habit pernicious in common media. "Their preliminary data suggests that in some groups, the more immigrants participate in activities organized within their cultural community, the more they also participate in activities that bring them in contact with other Canadians, like a community sports team." Since we can't draw a causal relationship from this information (Correction: 'Shouldn't' draw a causal relationship; we can draw anything we want to - except ponies. Their legs are oddly shaped) functionally all they are saying is: the more immigrants participate in activities the more likely they are to participate in activities. Crazy.

I run into this sort of disillusioning article every so often when I open the paper and forget to disable my standards; undeveloped information that hides an ugly agenda under the guise of innocent fact-finding (See: all psychological research about race from the 70's). I totally support research for research's sake; I am a philosopher for heaven's sake, the same group who seriously considers the dynamics of a universe whose only occupant is a bucket of water on a string. When confronted with a direct goal for our musings we tend to shriek and melt like a witch with a water allergy, but sometimes we have to take a step back and wonder what the research is saying about ourselves.

No comments: