Assessment and the Pity of Induction
In school mathematics, Desmos and Geogebra online software augment learning. The apps clarify thought; so that students don't bog down in minutia. They protect working memory from being overwhelmed with mechanics. WriteLab delivers the same benefits to writers.
Reading provides knowledge. Writing drives coherence and winnows error. It clarifies thought. Unclear structures muffle thoughts and frustrate readers. WriteLab reduces the struggle to communicate; yet, more importantly, it allows writers to focus on connecting content by maximizing their working memory. With it, writers gradually learn appropriate English usage. Whether editing Facebook posts, documents within Chrome, or online test responses, WriteLab has a purpose and a place.
The power of WriteLab tempts those who want to assess someone else's writing. Extending it appears as a mere engineering effort, but it's not. Reaching into other domains; in particular, content invokes induction; where finding solutions may not be complex, but complicated or unsolvable. Remember high school math; there are very few proofs by induction. Brute force cannot overcome logic.
Severely restricting a content domain may allow an enhanced WriteLab-type product to serve as an auto-grader, but the resulting reductionism weakens the very intent of writing. A variation of Close Reading has entered as the principle strategy at the heart of the SBAC/PARCC Common Core assessments. Limiting responses to just to statements found in a short text has little to do with education. It's the opposite! Thinking requires bringing external knowledge to bear on new information. Writing tests the strengths of the new connections. Modern close reading, as actually taught beginning in first grade, becomes a copy and paste approach to writing - merely a game.
Understanding the appeal of close reading, which works well with WriteLab, surfaces beliefs that strike the center of popular educational myths. First, close reading makes auto-grading feasible, reinforced by faith in technological progress. Second, close reading seems fair. Little or no content knowledge outside of the text is needed. Third, educators can close the achievement gap by limiting the value of the knowledge of higher performing students. Fourth, and most importantly, close reading allows the current individualized approach to English and the Social Sciences to continue. There is no national curriculum. Students don't read the same books which inherently limits shared content domains, which then only permits assessments of limp skills. Fifth, elementary schools rely on programs such as Accelerated Reader, where children make their selections, buttressed by the concepts of personalization and differentiation. The lack of a standard curriculum, not standards, creates no choice for assessment makers but to use neutral or self-contained texts, which close reading complements. The “benefits” of this closed circle accrue to all in education; except to students, who are kept from the richness of learning and the ability to show and share it, knowledge that allows exceptional writing, where a tool like WriteLab can benefit them deeply.
Perhaps an alternative exists where a moderately reduced content domain may not only evade the educational problems caused by isolated texts and close reading, as currently practiced, but also, the constraints of induction. For example, it would be reasonable for the SBAC Consortium to state one day that "this year, many SBAC reading questions will concern the settlement of the Western United States. Students with an understanding of the following text sets will be able to address the assessment more fully. Please realize that SBAC answers using materials outside of the text sets create more difficulties in grading, which is inherently difficult in the first place."
The fragmented United States polity and their beliefs in individuality and progress prevent a national curriculum providing the flesh on the bones of standards. However, a minuscule shared core may allay concerns. Close to three decades ago, when E. D. Hirsch originated the Core Knowledge curriculum, he voiced an interesting idea. The CK curriculum would comprise only 50% of class time, leaving room for local content (j.mp/EDHirsch50). An SBAC established 5% solution would allow richer, deeper assessment and, possibly, reduce the fade in NAEP US students experience from elementary through high school.
WriteLab would assist students throughout the learning and testing process. The time spent in summative testing would also serve as a learning experience. Whether or not, restricted-domain auto-grading circumvents the problems of induction, WriteLab truly serves as a lab for student learning whatever situation it which it resides.