even about justice and fairness, is to avoid the racial, avoid speaking of racism. So I choose not to. The conversation needs to happen. It hasn’t in writing assessment circles.
I’m mindful of Stephanie Kerschbaum’s (2014) work on the rhetoric of difference in the academy. I realize I could be engaging in what she calls “taxonomizing difference,” a theorizing that often “refuses to treat racial and ethnic categories as monolithic or governed by stereotypes by recognizing the variation within categories,” but the categories offered tend to be “relatively static referents” (p. 8). This denies the individual ways that students exist and interact in language and in classrooms. Or I could be engaging in what Kerschbaum calls “categorical redefinition,” which “focuses on producing more refined and careful interpretations within a specific category”(2014, p. 10), for example, my insistence throughout this book on seeing the Asian Pacific American students at Fresno State as primarily Hmong students. Both rhetorics of difference, according to Kerschbaum, can allow the researcher or teacher to place identifications and associate cultural and linguistic attributes to students instead of allowing any differences, and their nuances within supposed racial or ethnic categories, to emerge through actual interactions (2014, p. 9). So one dance I attempt in this book is to talk about race and racism in writing assessments without forgetting that every individual embodies their racial identity in unique linguistic and other ways. But there are patterns. We must not lose them in our attempt to acknowledge individuality.
More important, if you can see the way racism is one product of all writing assessments, then you can see the way biases against non-heterosexual orientations might be, or certain religious affiliations, or gender bias, or economic bias. The dynamics are similar. They are all dynamics of power, but they are not historically the same, and they are not just about bias or attitudes toward people. I’m not, however, suggesting that these dimensions of difference are equal in social weight or consequences, that the oppression experienced and felt by students who proclaim a Christian identity is on par with the issues that male African-American students on the same campus face, nor am I suggesting that any of these dimensions are separable. Of course, we cannot simply think of a student as one-dimensional, as her race, or her gender, or her sexual orientation, or her class upbringing. All these dimensions intersect and influence each other, creating individuals within groups who are as unique and different from each other within a racial formation as they are from those of other social groups. Kimberle Williams Crenshaw (1991) explains this phenomenon as “intersectionality,” a dynamic of oppression in which multiple structures intersect, such as issues of class, economics, culture, and race. So when I speak of race in this book, I’m thinking in localized terms, ones that assume local racial formations’ economic and other patterns as much as I’m thinking of racial structures. So while this creates diversity within locally diverse populations, we can still find patterns in those populations, as well as a few exceptions. The patterns come mostly from the structures people work in, and particular racial formations tend to be affected and moved by particular structures. This creates the racialized patterns. So I’m not interested in the exceptions, only the patterns. As a culture, we (the U.S.) focus too much on exceptions, often fooling ourselves into believing that because there are exceptions, the rule no longer exists or that it’s easily broken by anyone with enough willpower or hutzpah.
In popular culture and talk, race is often a synecdoche for a person’s physiognomy, heritage, culture, and language, even though these things cannot be known by knowing someone’s self-identified racial designation, or by their physical appearance, or some other marker of race. Race is also easily seen by most people as a construct that should not be held against a student, nor should it be used to judge the merits of their labors, yet few deny that most large-scale writing assessments are racist, or at least reveal different performance patterns that are detected when results are disaggregated by racial formation. Many have already discussed the negative effects of various writing assessments on students of color (Fox, 1999; Inoue & Poe, 2012a, 2012b; Soliday, 2002; Sternglass, 1997; White & Thomas, 1981). Others have investigated the effects of a variety of large-scale tests on students of color (Hong & Youngs, 2008; Jencks & Phillips, 1998; Madaus & Clarke, 2001; Orfield & Kornhaber, 2001; Plata, 1995). We can find racist effects in just about every writing program in the country. We live in a racist society, one that recreates well-known, well-understood, racial hierarchies in populations based on things like judgments of student writing that use a local Standardized Edited American English (SEAE)1 with populations of people who do not use that discourse on a daily basis—judging apples by the standards of oranges. Racism has always been a part of writing assessment at all levels.
Some may argue that I’m painting this picture of racism in writing assessments, in writing classroom assessments particularly, too broadly. I’m lumping the accurate judgment of performance of say some Blacks who do not perform well into the same category of writing assessments as those assessments that may exhibit cultural, linguistic, or racial bias in the judgments or decisions made. To put it bluntly, the argument is that sometimes students do not write well, and they should be evaluated accordingly, and sometimes those who do not write well will be Black or Latino or multilingual. Just because a writing assessment produces patterns of failure or low performance by students of color who participate in it doesn’t mean the assessment is racist. This is an important argument. I do not argue to let students slide academically because they happen to be by luck of birth a student of color.
On the other hand, I see a problem with this argument. Why do more Blacks, Latinos, and multilingual students relatively speaking perform worse on writing assessments than their white peers in writing classrooms? At Fresno State, for instance, between 2009-2012, the average failure rate for Blacks in the first-year writing program was 17.46%, while the average failure rate for whites for the same years was 7.3% (Inoue, 2014b, p. 339). Whites have the lowest failure rates of all racial formations, and this is after the program revised itself completely in part to address such issues. That is, these are better numbers than in the years before. I realize that there are many ways to fail a writing class beyond being judged to write poorly, but these internally consistently higher numbers that are consistent with other writing programs suggest more, suggest that we cannot let such numbers pass us by just because we can assume that teachers are not biased.
I’m not saying we assume bias or prejudice. I’m saying let’s assume there is no bias, no prejudice. Now, how do we read those numbers? What plausible assumptions can we make that help us make sense of these data, what rival hypotheses can be made? Do we assume that more Blacks, Latinos/as, and Asians at Fresno State are lazier or worse writers than their white peers? Is it the case that on average Blacks, Latinos/as, and Asians at Fresno State simply do not write as well as their white peers, that there is some inherent or cultural problem with the way these racial formations write? Or could it be that the judgments made on all writing are biased toward a discourse that privileges whites consistently because it is a discourse of whiteness? Could the writing assessment ecologies be racist?
I am mindful of the concern in the psychometric literature that mean scores (like those I cite above) do not necessarily constitute test bias (Jensen 1976; Reynolds, 1982a, 1982b; Thorndike, 1971). I’m not concerned, however, with test bias in the psychometric sense, which amounts either to intentional bias on the part of teachers, or a disregard for actual differences that do or do not exist among populations of people (Inoue & Poe, 2012b, p. 352; Reynolds, 1982a, p. 213). In one sense, I’m concerned with writing assessment as a much larger thing, as an ecology that is more than a test or an essay or a portfolio or a grade or a rubric. I’m concerned with what might broadly be called fairness in the ecology , which is a measure of its sustainability. In an important article on how legal definitions of disparate impact can be used to understand assessment consequences, Poe, Elliot, Cogan, & Nurudeen explain fairness: “the Standards for Educational and Psychological Testing provide four principal ways in which the term fairness is used: lack of bias; equitable treatment in the testing process; equality in testing outcomes; and the opportunity to learn” (2014, p. 592). I’m most concerned with the second, third, and fourth items, but the first is also a concern, only not in terms of individuals but bias in the ecology. For Poe and her colleagues, they see much of fairness in assessments hinging on whether an assessment has disparate impact, which can be understood through an assessment’s methods. They explain:
the