ABSTRACT
Myths and Assumptions Surrounding Test Technologies A sociological discourse of assessment, as well as identifying functions concerned with perpetuating social disparity, also addresses the question of why those functions rarely appear as the focus of public awareness and debate. For example, George Madaus (1994:79, citing Winner) points out that most Americans do not enquire whether a test may produce a set of disadvantages or inequities along with its professed advantages. Most people are simply not aware of the biases and assumptions of technical elites that underpin test content and processes. In Chapter 3 of this book, writing with Cathy Horn, he makes the point that testing is a social technology with social consequences, yet people no more concern themselves with how a technology like testing works than they do how light bulbs or telephones work. From a similar perspective, Broadfoot (1996:232) argues that the emphasis in public debate on the ‘scientific’ nature of assessment has consistently disguised the values and power relations, the interpretative and the idiosyncratic in assessment practices. Darling-Hammond provides a useful example of such disguised cultural values and power relations in the manipulation of outcomes in the early development of IQ testing. These tests threw up inequalities in performances favouring urban over rural groups, higher over lower wealth, English speakers over immigrants, whites over blacks, and so on. Quite simply, these differences were taken to reflect and confirm what every intelligent person ‘knew’; that these groups were inherently unequal in their mental capacities. However, when girls outperformed boys on a particular test, this clearly reflected a ‘flaw’, and future test items were ‘corrected’ to create parity of results for boys. Differences of race, colour or social class did not warrant revision, though, as the validity of those findings seemed patently obvious (Darling-Hammond 1994:10). It must not be assumed, however, that the kinds of biased assumptions and manipulation of items and scores that we see in early mental testing have been eradicated with more sophisticated understandings and techniques. Mark LaCellePeterson in Chapter 2 gives us some prime examples from the testing of language minorities. His account includes the report that newly developed ‘performance’ assessments (see Introduction to Part I above) of early language development were going to be rescored because children in bilingual programmes scored ‘too high’ on them. Of course, as I stated in my opening Introduction, test developers are concerned to monitor for bias in their methods and to maintain confidence in systems of assessment and outcomes. However, through Part V of this book in
particular, the case is put that there is no place to stand outside of culture from which particular groups can set standards and claim to act as disinterested judge (Torrance, see Chapter 9).
