<p>Kryptonsa, I was being flip. But still, it would be very difficult for a great academic writer to get a 3.0. If someone scores a 4.0 or under, then that’s probably the only time AW DOES count. </p>
<p>Most professors haven’t seen the AW test. They haven’t read ETS’s examples of essays in the different score tiers, so they don’t know what the section tests. If they were given those essays and told to rank them in order of skill, they would grade them exactly as ETS did. They might even start caring about those scores. But you’re right. Right now, they give AW scores only a glance.</p>
<p>The problem for me is the recent change in ETS’s method of scoring the essays: they use one human grader and a computer program. If the two agree, then that’s your score. Before that change, they used two human graders, with a third called in if the scores varied too widely. Anyone who knows even a little about computers knows that they cannot measure logic, critical thinking, and other intangibles; they can only pick up signs that those elements might be there. The human grader is more reliable, but what happens if that grader is biased against your style of argument – or even against the argument itself? The two-person system protected test-takers from this kind of bias.</p>
<p>But few programs care about the AW not because they think it’s inaccurate but because they care much more about the rest of the application. In the humanities and social sciences, applicants have to send writing samples relevant to their field, a much better indicator of their analytical writing ability. In the sciences, lab research outweighs writing ability. If you’ve published a paper, then it is also assumed that you can write.</p>
<p>That said, probably the biggest general complaint among science professors is that too many of their graduate students cannot write. When I say “cannot write,” I’m not talking about the ability to write sentences and paragraphs, which every graduate student can do, but instead about the ability to write clear, logical prose that conveys complex thought and that progresses smoothly from one idea to the next. This is what the AW purports to measure. Does it? Well, yes, at least in theory. Test-takers run into problems if they think it measures writing skills and therefore don’t prepare for it. As someone else said, writing well is only part of it; you must be able to analyze an argument and present a well-written, nuanced short essay. The section tests logic, critical thinking, progression of ideas, strength of conclusions, etc.-- everything your advisor will value. If your PI/advisor is rewriting huge portions of your papers (and you should be reading the finished product to know this), then you aren’t writing well enough. Professors gripe all the time about doing work that their graduate students should have done, and that includes writing.</p>