@gagne , the New Yorker test applies to all New York schools that are fully funded. It also applies to Hunter and Brooklyn College because those schools are selective, even though not fully funded.
The premise of the test is what Adcom's regularly say. 80% of applicants to fully funded program are rejected after reading one page because the applicant clearly is unaware of conventions in literary fiction.
You take the last 6 fiction pieces from the New Yorker Magazine. You mix in 2 of the applicant's stories. You ask anyone (no matter how illiterate) to seek the odd man out. If they identify the applicants 2 stories, then the applicant will be rejected after one page. The test obviously won't work if the writing samples are experimental.
Take for example the links @koechophe posts. If you mix in his stories with New Yorker stories, any reader will instantly flag the two stories as devoting 50% of words to explanations. The New Yorker test will be failed. Because of this test, 80% of applicants don't have to wait for school decisions. An applicant can even endeavor to adjust their writing sample to something that would pass the test for next year.
People often complain that the test is not 100% reliable. That is true of any test, but these detractors are also the first ones to admit they know of no more reliable test for predicting admission. (Their arguement is like saying GPA does not perfectly predict which undergrad will accept you.)
