Assessing Assessment
Cynthia Haynes once wrote that in the teaching of argument, everything is up for argument except the teaching of argument. One might say something similar about assessment.
As a WAC director, I hear a great deal about assessment. It’s a major topic of discussion at the WPA conference. It is a major topic of discussion at IWAC. The journals show numerous pieces devoted to assessment. Discuss new media and writing and the first question you get is, “how do you assess that?” Assessment is the great mystery for WPAs. It is magical (it supposedly can save or destroy programs on its own). It is a safety net (good assessment supposedly will keep programs from being cut), it is an obligation (the mantra of some “mandate” from some place else). It provides standards (it supposedly keeps us following goals or set outcomes). WPAs are obsessed with assessment. They assess their programs. They assess student work. They view the field through the lens of assessment.
The topic of assessment, however, is seldom put under assessment. Take assessment methodology as example. There is a widespread assumption that quantitative data analysis is the method of assessment. Such methodology might include gathering data about GPA correlation, might be based on rubric matching, might involve communal reading of student work, and might involve the collection and communal reading of student portfolios (with matching to a rubric following). This methodology depends on keywords: reliability, validity, measurement. These keywords provide a specific type of focus or lens. It is, though, a limited focus.
Typically, this work is meant to demonstrate to a supposedly hostile administration the value of the instruction being done. The work provides value to the WPA her/himself who often feels under the gun, unappreciated, and undervalued within the academic infrastructure for one reason or another. The work makes writing instruction slightly more “scientific” (we are gathering “data” after all) and thus on equal par with other university work. Thus, the keywords noted above. They are the keywords of science studies (to some extent). Assessment is posed as science.
One problem, though, is that in this type of conversation, assessment is treated as a given or a Barthesian mythology. The assumption being that assessment (at least of this sort) is “just something one does.” Or those involved in assessment follow the Sammy Hagar mantra and change “there’s only one way to rock” to “there’s only one way to assess.” Another problem is that the gathering of such data (often done through reading student work and matching it to a rubric, student surveys, faculty surveys, GPA correlations) demonstrates causality. Indeed, causality - doing X caused one to write better or worse - is almost impossible to prove (see ANT).
Assessment depends too strongly on the concept of proof. Proof is a topos. Our values and world views depend on communal topoi (or we would have, as McLuhan noted, psychosis). As one WPA-L listserv comment notes, “gain” is the goal of this world view. Assessment should “prove” something called “gain” which tends to mean “improvement” or “skill” acquisition. Again, see ANT.
I write this because I share the contradictory spaces of WPA and cynic regarding assessment. To put it another way: While I may run a major writing program, and while I am always interested in understanding how our program functions, I am cynical of the assessment practices familiar to our field. They are repetitive (many reach the same conclusions), involve tedious work (laborious collections and reading of papers that are solved by paying readers since no one person can ever read all this work), never fully demonstrate proof (even as they claim otherwise) and in the end are done mostly because of programmatic insecurity (”we’re good, we’re really good!”). I, too, though am engaging in assessment. Our program is assessing itself. But we aren’t following this tradition I briefly outline above. Instead, we are following a practice I am calling “Networked Assessment,” a practice that treats the program as a network and that poses assessment as the network practice of tracing an account (see Latour). We are not out to prove value. We are working to trace the relationships of our program in order to better understand how our program works (the way Latour does with law in The Making of Law or with the failed Aramis project in Aramis). Value could be found in such a tracing, and likely will be found as relationships are discovered within the network, but it is not the goal. And in such a tracing, validity, reliability, and measurement are not the keywords that provide a lens. The main keyword is description.
I’ve given a couple of talks on this subject and have had a piece that outlines the concept and approach recently accepted for publication. I want to write more on this topic in the next few days, but for now, this post introduces the challenge to assess the concept of assessing. Who watches the Watchmen, and who assesses the assessors? Let’s trace the assessors. Let’s provide an account of their work and relationships. Let’s put assessment under assessment.







