Where Did Writing Tests Come From?
The modern standardized writing test has had a major effect on how writing is taught and how people write. When I directed writing programs at both Tulane University and the Massachusetts Institute Technology, a major effort of first-year writing classes was to deprogram students from writing the five-paragraph essay that had helped get them into these elite institutions. Why, then, is this form, which does not exist in the real world, so ubiquitous in classrooms?
First, that which is tested is that which is taught. The prominence of the five-paragraph essay derives from over a century long history of attempts to assess the writing ability of students. Writing, like all forms of communication, is a complex and multidimensional activity. Compared to assessing writing, assessing skills such as mathematics are relatively straightforward. With writing, however, it begins with the very basic question of what do you want to measure. And the history of writing assessment illustrates that this question has been very difficult to answer.
Like much in American education, some of the trends in writing assessment began at Harvard. In 1874, Harvard added an entrance examination in English Composition. The test simply required students to summarize plots from set literary texts. Readers determined grades solely on the accuracy of the summary and correctness in spelling, grammar, and punctuation. The test provided students no opportunity to interpret or analyze the text.
In contrast, when the College Entrance Examination Board offered its first college admission test in 1900, this test, although still based on literary texts, asked students to write at least two pages of analysis and argument. The first prompt asked students to compare the characters of Achilles and Hector in The Iliad. Moreover, students had a total of four hours, two hours in the morning and two hours in the afternoon, to write. Cognitively, it was a much more difficult task and one that was much closer to the types of writing students do in college. Moreover, students had time to plan, write, revise, and edit their essay.
Yet since the 1970’s, the College Board and most other testing organization have relied on a type of writing test known as the Timed Impromptu. The basic features of this kind of test are that students are given a short amount of time to write, usually between 20-45 minutes, on a general topic that, at least in theory, is equally assessable to everyone, such as one of the prompts for the then new SAT Writing Test in 2005, “Should schools help students understand moral choices and social issues?”
The use of the timed impromptu is largely motivated by its being inexpensive in both time and money and by its reliability. Testers had long complained that readers scoring essays could not achieve the kind of reliability that was produced by their beloved multiple choice tests. Readers were able to agree on scores of timed impromptus much more often then they could when reading longer essays based on texts and when students had time to plan and revise.
There is just one major problem with the timed-impromptu. It bears very little relevance to the kind of writing people do in real life. Everyone has had situations at work and in school where they had to write quickly on demand. But in these situations, be it an economics essay question or an intra-office memorandum, the writer is always familiar with the subject matter and context. No one has ever received an email from their boss, asking them to respond within 25 minutes to the assertion that failure is necessary for success.