Or is it the fault of the students who demand a certain kind of teaching? […] Is it perhaps the fault of the teaching institutions, which do not provide any kind of teacher training in TOEFL preparation? […] Or is training in teaching test preparation the responsibility of the textbook writers and publishers? (p. 335) Hamp-Lyons’ (1998) questions remind me of the social nature of morality. The point of her questions, as I understand them, is not to apportion ultimate blame (“yes, it’s the students’ fault”) but to point up the fact that adjudicating on moral issues is a highly complex process in which many individuals and institutions have a stake. Standardized tests are social phenomena par excellence; any consideration of their moral significance must begin from this starting point. 2Of course, I am sidestepping the fact that with a so-called communicative test it is still necessary to define what knowledge of language is and to ignore the fact that an examination virtually by definition cannot involve genuine communication and therefore will always be only an indirect and artificial indication of the candidate’s “true” ability, however ability is defined. The Morality of Testing and Assessment 72 For example, much as I loathe the whole business of tests, when it comes to my own students I can understand why they would want to have extra preparation. Although at one level such an approach demolishes both the illusion of the snapshot-of-ability principle and the principle of equality and fairness, on the other hand, the moral importance of relation creeps in. It is never the case, and I would argue that it never should be, that a teacher’s own students are not more important than some other students in another state or country. We want the best for our own students, even if in general moral terms it could give them an “unfair” advantage. This is a classic instance of the way in which individual circumstances and specific relations color our approach to moral dilemmas. This brings me to a final point, which Noddings (1984) raised in her discussion of assessment. Contrary to received wisdom regarding the preferability of local, teacherdeveloped forms of assessment over mass standardized testing, Noddings wrote that she
is “convinced…that grading—summative evaluation of any kind—should not be done by teachers. If it must be done, it should be done by external examiners, persons hired to look at students as objects. Then teacher and students would be recognized as together in the battle against ignorance” (p. 195). Despite my instinctive and growing distaste for standardized tests of all kinds, I find Noddings’ argument curiously persuasive because, like her, I believe that our prime duty as teachers is to focus on the learning of our own students. Turning the problem of evaluation over to outsiders moves it from the immediate, local teacher-student relation, nd spares that relation the “grinding” experience mentioned earlier in which the teacher switches caps from advocate to judge. Noddings’ (1984) suggestion does not justify indiscriminate use of testing and it does not offer any justification of current testing practices or excuse test makers from an obligation to continually rethink the format and nature of their tests. It does, however, remind us that we are dealing with issues of immense moral complexity, in which unequivocal good and bad, right and wrong, are terribly hard to pin down.

Continue Reading

One candidate from a western Eu-ropean country had an English mother, and her entire application confirmed her own categorization of herself as “virtually bilingual,” yet her TOEFL score was a mere 597, which, the computer-generated form from the university’s Office of International Admissions told us, indicated that she “may need supplementary English training.” A doctoral candidate from an African country in which English is widely spoken had attended an Englishlanguage university for his undergraduate and master’s degrees and had sterling references and published academic work, yet his scores on the TOEFL and Graduate Record Examination were both abysmally low. What should we do in such situations? Both candidates were admitted, but in each case the test scores complicated the decision rather than making it easier; in the case of scores on the Graduate Record Examination, for example, we are required by the university’s graduate school to obtain an official exception for any candidate who does not score the required minimum in this test. 71 Values in English Language Teaching Another moral paradox is the disjuncture between testing and current pedagogical practice in language teaching (Hamp-Lyons, 1998). The communicative model that, in various forms, is widely used across the world encourages students to engage in meaningful interaction using whatever linguistic means they have at their disposal; it specifically downplays the importance of grammatical accuracy over communicative effectiveness. Although some tests have made more or less successful attempts to integrate communicative competence in their evaluations of students (the Cambridge suite of examinations and the ACTFL’s [American Council for the Teaching of Foreign Languages] proficiency guidelines for foreign languages [Byrnes & Canale, 1987] come to mind; see also Powers, Schedl, & Wilson, 1999), the TOEFL has notably lagged behind in this regard, and many other tests still focus narrowly on grammar and vocabulary.2 One consequence of this disjuncture is the so-called washback effect: the ways in which test format affects teaching. The theory behind the TOEFL, like any test of its kind, is that it is a snapshot of a candidate’s language ability at a given moment in time; thus, it should not be possible to improve one’s performance other than by more study of the language. In reality, of course, TOEFL preparation courses and programs abound. Liz Hamp-Lyons (1998) offered a very thoughtful analysis of some of the ethical (what I would call moral) issues that arise from the “powerful” (p. 331) washback effect of such an influential test as the TOEFL. Drawing on the work of Mehrens and Kaminsky (1989) and Popham (1991), Hamp-Lyons posed the question of what constitutes “ethical test preparation” (p. 334) and argued that existing materials (and hence a great deal of existing test preparation programs across the world) are “educationally indefensible (boosting scores without mastery) and of dubious ethicality (coaching merely for score gain)” (p. 334). She went on to ask a series of provocative and important questions, all of which have a strong moral dimension: Can a test be blamed for the ways in which some teachers teach towards it? […]

Continue Reading

This, combined with the strict and ritualistic way in which standardized tests are conducted, gives the knowledge they enshrine a solemn, almost sacred significance. The political, “interested” nature of knowledge featured in chapter 3 is a powerful component here too. I argue that there is deep moral meaning in such an approach to knowledge: By reducing learners to recipients of knowledge rather than creators of it, one is also reducing their capacity for moral agency. There is also a question of honesty here. Because of the veil of objectivity behind which they hide, standardized tests ride roughshod over the unavoidable difficulties of matching score with actual ability. The final score is presented (and in the overwhelming majority of cases is also treated) as an objective measure: The uncertainties and ambiguities that attend test development, and the myriad psychological factors that affect a candidate’s performance on a given day, are invisible. Furthermore, because of the physical and administrative distance between the testers and those tested, appeals are difficult, if not impossible. A teacher might possibly be inclined to be lenient on a student whose grandfather died a few days before the exam, or to give a student who has difficulty writing an extra minute or two at the end of a test. A standardized test can offer neither of these possibilities or anything like them. What is missing here is relation: The human relation between tester and testee, which exists when teachers prepare tests, and which informed the whole of the previous section, is entirely absent in the standardized test. By this account, the moral contours of the test are quite different. The educational process is suddenly deprived of its deepest and most meaningful component. This feature is underlined even more in the current shift to computerized testing in the TOEFL and many other common tests. The impersonal nature of such tests, and the impossibility of our understanding the human dimension of the test-taking experience of any specific individual, makes it very difficult for the consumers of test score information to know how to interpret them. As language professionals, we know the complexities I have been discussing here; as a result, reading the scores is very much a matter of interpretation rather than a simple acknowledgment of a score. Just 2 days ago I was reviewing some late admissions for our own master’s and doctoral programs.

Continue Reading