Sexing the Database, page 3
The policies of institutions like UCSF might lead to problems interpreting data on how common intersexuality and transsexuality are -- as well as what kinds of medical conditions one might want to watch for in these populations. Transsexual women, for example, are subject to breast cancer from taking estrogen; they are also, by virtue of having been born in male bodies, subject to dangers from prostate cancer. How can these diagnostic possibilities be captured in an M or an F? The answer is that they can't. Not only are statistical data disrupted by clinging to categories like M and F, but the lives of individuals could be put at greater risk because doctors consulting the database won't be getting the full story on what conditions their patients are vulnerable to.
Staff at UCSF make up for the inadequacies of their databases in the good-hearted, casual, not-quite-on-the-record ways people in large institutions often do. They create what Cox dubs "a note in the medical record" and Cabello calls "a notation only." What both administrators mean is that when a patient doesn't fit the accepted parameters in a large database (and that would be most patients, some of the time), caregivers will give more complete and helpful details about the patient's condition in a note somewhere. It may quite literally be a post-it note in the patient's physical records file, or it may be in a "notes" field of the database. At UCSF, such notes are what allow people outside the M or F categories to be given appropriate treatment. According to Cox, doctors will leave a transsexual's gender category unchanged in the patient databases, but will make a note about transsexuality somewhere in the patient's file. Unfortunately, such notes are generally what go first when records are passed from one database to another -- or from one filing system to another. They are the crucial but ephemeral bits of data that get lost when you translate everyone's bodies into two classifications known as M and F.
What Kind of Female Are You?
WashingtonPost.com could care less if you're a transsexual woman, a natural-born woman, or an intersexual who calls herself female on Tuesdays only. All they want to know is whether they can sell you to advertisers based on your gender identification the moment you log on to their site and ask to read the news. A representative from WashingtonPost.com said the sole reason for accumulating gender information about their readers is "advertisers."
Gender-targeted advertising is as old as advertising itself -- just take a gander at some of the vibrator ads from early-twentieth century issues of Woman's Home Companion. The advertising industry has used demographics -- which include data points on a target population's gender, age, education and income -- to tailor their ads in fairly general ways. Demographic data is what drives toy companies to buy commercial time during after-school hours. Demographics also make it obvious that pharmaceutical companies should place ads in magazines that doctors read.
But recently newspapers and other businesses which earn money through advertising have opted to deliver their readers to advertisers with for more than traditional demographic data attached. Gathering reams of data on your readers, clients or website visitors reached the height of its popularity in the late 1990s, when "psychographics" began to supplant "demographics" in the advertising industry. A psychographic profile of a consumer is sensitive to the consumer's values, beliefs and attitudes. Gender is only one small piece of the personality pie when advertisers make suppositions about what a person's psychographic might be: for example, if you're female, college-educated and Republican, your psychographic might reveal that you tend to shop at The Gap for "office casual" clothing. Gender is certainly part of the psychographic data used for advertising, but it matters far less than other factors when advertisers try to appeal to what they imagine a consumer's "attitude" might be.
Just as the advertising industry was reaching a psychographic frenzy, encouraging massive amounts of customer data-gathering and data-trading on-line, the tech bubble burst. At the same time, privacy advocates like Simson Garfinkle, Jeffrey Rosen and Declan McCullaugh were publishing digital muckraking articles and books about all the creepy ways advertising companies were tracking people on-line. These criticisms, coupled with the sagging economy, cast a pall on psychographics. Companies still want psychographic data, but they can't be as open about gathering it as they once were. It's likely that WashingtonPost.com's data on gender is far less valuable than information it has databased on readers' dining habits, the number of vacations they take per year, and whether they plan to send their children to college. And yet casual readers of WashingtonPost.com are perhaps more comfortable giving simple gender information than they are spilling their guts about their whole lives on a form. But gathering gender information by itself just isn't useful to advertisers anymore. What good is it to know that fifty percent of your readers are women if you don't know what kinds of jobs they have and how they feel about drinking alcohol?
| Next: Data Points of the Future | Pages: 1 2 3 4 |
 |
|