You could say I stumbled upon this post by accident. Although I grew up in the digital age, I have a love-hate relationship with all things computerized. In high school, while all my classmates were busy typing away in the computer lab, I could usually be found curled up in the hallway, furiously scribbling in my five-subject notebook. Despite the fact that I’ve been proficient on word processors since I was about nine years old, it took me until my sophomore year of college to teach my brain to work in front of a computer screen. A notebook and pencil are still my preferred writing tools, and I almost always use them, at least to get started. Nevertheless, I completely understand why so many people prefer computerized composition, and I don’t deny that word processors are incredibly useful in the writing classroom. Despite my own personal preferences, I can’t imagine teaching a writing class without allowing my students the option of drafting and editing on a computer. It was my opposing feelings about computer-aided composition that initially attracted me to the article “On Writing Tests, Computers Slowly Making Mark,” which turned up in my feed reader despite the fact that it’s slightly dated. Published in Education Week in February 2007, the article discusses the possibility of the writing portion of the National Assessment of Educational Progress (NAEP) becoming computerized. Once I began reading, however, I was drawn in not by the debate over the merits of computerized writing assessment, but by the way that debate illuminated certain issues that are relevant to standardized writing assessment in general, regardless of how it’s administered.

          One of the concerns the article raises about computerized writing assessment is that schools vary in the number of computers they have at their disposal. For that reason, some schools would need a much bigger window of time to administer the test to all their students. I know what you’re thinking. So what? Who cares? Let them do it over two or three weeks instead of a few days. What’s the difference? Well, as the article puts it, “that strategy can raise concerns about test security.” Hmm. There’s something very wrong with that statement, besides the fact that it’s reminiscent of a press release from the Pentagon. Why should the test have to be kept a secret? My interpretation of the reason is this: if schools are allowed to administer the tests over a longer period of time, student A, who takes the test on day 1, could potentially reveal the prompt to student B, who won’t take the test until day 7. Thus, student B would be left with six precious days to plan his or her response. My question is, why do people spend so much time thinking about the security of test, without ever bothering to ask themselves why a writing test has to be secure to begin with? I’ve never heard of any professional writer who can sit down and produce his or her absolute best work in only a few hours, without the benefit of being able to think about the topic, talk to others about it, or research it. The very idea is laughable. In fact, I don’t think many writing teachers would encourage students to do that kind of writing if it wasn’t mandated by high-stakes testing. If the real goal is to obtain an accurate reflection of students’ writing ability, and if we insist on doing so without allowing them to choose their own topics, we should at least be willing to pass out the test prompt a few weeks beforehand! We should be talking about it in class, and researching it in the computer lab! We should be watching movies and singing songs about it! We should be doing whatever will help our students gain as thorough an understanding of the topic as possible before sitting down to take the test. The issue is not how to make computerized testing more secure. The issue is the fact that test security is a concern to begin with. Whether it’s computerized or not is irrelevant.

          The concern over security fits quite nicely with another aspect of computerized writing assessment addressed in the article. If a test is computerized, should students be allowed to use the spell-check and thesaurus programs that are standard in most word processors? The board that was developing the computerized version of the NAEP writing test at the time of the article’s publication believed they should, because “those word-processing devices have become so prevalent in schools, and society at large, that it does not make sense to ask students to work without them.”  Director Rosanne Cook goes on to explain that, “it [word processing without the benefit of editing tools] creates a very artificial environment for the writing test that is not what they’re accustomed to.”

Now wait just a minute. If the claim is that it’s unfair to ask students to produce “good” writing under artificial circumstances, I couldn’t agree more. But what in the world would make anyone think that artificiality isn’t already an issue? As I explained earlier in this post, the conditions under which standardized writing assessments are typically administered are already structured in a very artificial manner. Nothing about them resembles the way real, non-test writing takes place. Furthermore, it could easily be argued that if students are familiar with being timed as they write to a prompt, it’s probably because of the testing virus that has infected their schools, making it illogical to argue that test writing isn’t artificial because it’s also practiced in classrooms. This is not a problem that could occur if a computerized writing assessment isn’t structured properly. It’s a problem that already exists, and will continue to exist as long as standardized writing tests have a place in our educational system. 

On Writing Tests, Computers Slowly Making Mark
Sean Cavanagh
Education Week
Vol. 26, Issue 23, p.10
February 14, 2007
Full Article