Last June there was a ruckus when an eagle-eyed parent named Jeanne Heifetz noticed that literary excerpts on the Regents’ exam her stepdaughter brought home had been altered, apparently with the intention of removing anything that might possibly give offense to anyone (including all references to religion), in the process seriously altering the meaning of the passages. After prolonged and well-deserved ridicule, the Department of Education caved and promised that changes would be made. The story faded from view.
Now Michael Winerip reveals in today’s NY Times that (as might have been expected by anyone with a healthy degree of cynicism) nothing has in fact changed.
Ms. Heifetz, bless her, recently got a look at August’s English exam. In new guidelines, the state promised complete paragraphs with no deletions, but an excerpt from Kafka (on the importance of literature) changes his words and removes the middle of a paragraph without using ellipses, in the process deleting mentions of God and suicide.
The new state guidelines promised not to sanitize, but a passage on people’s conception of time from Aldous Huxley (a product of England’s colonial era) deletes the paragraphs on how unpunctual “the Oriental” is.
But the saddest example of how standardized testing is lowering academic standards (as a recent national study by Arizona State University reports) can be seen in the way New York officials butchered an excerpt from a PBS documentary on the influenza epidemic of 1918.
Like any good historical work, the documentary on this epidemic, which killed half a million Americans, included numerous interviews with historians, novelists, medical experts and survivors, and quoted primary sources of the era. But the three-page passage read out loud to students on the state exam is edited to make it appear that there is only one speaker.
Though the new guidelines promised to identify the authors of any excerpts, the state does not identify the documentary’s author, Ken Chowder. It does identify the narrator, although — oops! — incorrectly: the narrator was Linda Hunt, not David McCullough. As Ms. Heifetz says, any student who melded the words of a dozen people into one and then misidentified the narrator would surely be flunked.
The state version cuts out the passages with the most harrowing and moving accounts of the epidemic, as when children played on piles of coffins stacked outside an undertaker’s home. It removes virtually all references to government officials’ mishandling the epidemic. It deletes the references to religious leaders like Billy Sunday, who promised that God would protect the virtuous, even as worshipers dropped dead at his services.
Furthermore, “Ms. Heifetz believes that one test question based on the influenza reading has three correct answers”—and the professor featured in the documentary agrees:
To get a second opinion on Question 2, I tracked down Dr. Alfred Crosby, a retired University of Texas professor who was featured in the PBS documentary and has written the book “America’s Forgotten Pandemic.” I sent him a copy of the state’s sanitized excerpt and the multiple-choice questions. Dr. Crosby loves history’s complexity and was offended by the state’s single-speaker vision of the past.
He believes all three answers to Question 2 were implied in the state excerpt and said that if he were marked wrong for responding with Answers 2 or 3, he’d be angry. “That’s the problem,” he said, “with a multiple-choice test.”
Visit the National Coalition Against Censorship site for more information on this and other stories, and suggestions on What You Can Do.
Argh! The forces of politkorreknost’ have done it again!
This strikes me as an absurd complaint. If the purpose of the passages is to provide something for students to interpret and answer questions about, as I suppose, then it seems to me that the test makers should be entitled to alter the text any way that they find useful for their purposes, provided that they have permission of the copyright owner to make a derivative work, and they attribute the text to the authors and provide an annotation (once and for all at the top of the test) that content may have been altered.
(“Ingredients: Peanuts. Warning: May contain peanuts.”)
Yeah, this strikes me as extremely misguided. Written passages are present in a test not to educate the reader, but rather to provide something for them to respond to and answer questions about. I am pretty sure that editing sections out is completely standard procedure in test design.
We actually discussed this issue in my high school A P English class. (I notice that the news story this old post is about was closer in time to that in-class discussion than to today.) One of the three essay questions on the Advanced Placement Literature and Composition test always involves (assuming it has not changed, and I can think of no reason why it would have) reading a short story and then composing a response to a question about the story. We practiced one of these essays in class, with a story and question taken from an old A P test, and I noticed that the printed text was actually an abridged version of a longer story that I was familiar with. I asked the teacher what we could do if a situation like this arose on the test, and she admitted that she had never thought about it before and did not know. We talked about the issue for a little while, and the consensus (ill informed as it was) was that if you were careful, you could probably bring in information about the parts of the story that had been edited out, saying something like: “In another part of the story, which was not included in the presentation of the question,….”
When I actually took the A P test, I again recognized that the short story question was using a condensed version of a longer story that I had read. However, I did not actually remember anything else that happened in the parts that had been excised, rendering the issue moot.
The ill-informed consensus approach Brett’s class settled on necessarily assumes that the grader: a) has actually read the full story; b) will be guilt-tripped into finding it and reading it (in the middle of a poorly-paid exercise where only a few moments are typically spent grading each essay); or c) will be guilt-tripped into taking the test-taker’s word as to what the excluded parts of the story say. None of those assumptions seem safe to me. My kids have been more heavily and routinely exposed to “DBQ’s,” as the jargon has it (document-based question, I think?) than I recall being in comparable grades, so by the time they take a Regents or AP test it seems likely that the issue of how much of your own happenstance prior knowledge not contained in the specific text put before you are you supposed or not supposed to draw on in writing your answer is one students in their generation will have dealt with before, and thus have a standard or default approach to. At least if they are the sort of student likely to remember anything relevant that isn’t explicitly contained in the text they are reading as part of the exam …
@J.W. Brewer: From what I remember, the handling of the document-based questions (DBQs) was peculiarly different between the A P United States History and A P European History examinations. The DBQ was the last essay question on each of the exams (and counted for the most, being worth 15 points, instead of the 9 for other A P exam essays). However, while on the U. S. History test, you were required to use external knowledge (not just what was presented in the documents) in your essay to get full credit, on the European History DBQ, you did not need to draw on an facts beyond what was given in the documents. While doing practice DBQs in class, I discovered that the different in approach meant that the subject matter of the European DBQs tended to be quite a bit more obscure, since there was no expectation that you would know much of anything about a specific question topic of time.
That seems plausible, given that most American students know more or less the same things about American history, but coverage of European history is invariably going to be spotty.
There are two separate issues here. I agree that it is reasonable (though distasteful to those of us who care about literature) to present simplified extracts for test purposes, but setting that aside, there is surely no excuse for expecting unique correct answers when the text as presented doesn’t uniquely support such an answer.