8. The assessment process should involve multiple perspectives and sources of data.
Perfect assessments and perfect assessors do not exist. Every person involved in assessment is limited in his or her interpretation of the teaching and learning of reading and writing. Similarly, each text and each assessment procedure has its own limitations and biases. Although we cannot totally eliminate these biases and limitations from people or tests, we can try to ensure that they are held in balance and that all stakeholders are made aware of them. The more consequential the decision, the more important it is to seek diverse perspectives and independent sources of data. For example, decisions about placement in or eligibility for specialized programs have a profound influence on a student’s life and learning. Such decisions are simply too important to make on the basis of a single measure, evaluation tool, or perspective.
The need for multiple indicators is particularly important in assessing reading and writing because of the complex nature of literacy and its acquisition (see standard 5). A single measure is likely to be misleading or erroneous for individuals or groups. For example, timed essay tests of writing can significantly underpredict the ability of English-language learners to write under natural conditions, and instructional decisions made on the basis of results on such tests will thus impede their educational progress. Multiple sources of data, on the other hand, can allow for triangulation in problem solving. Sources of data can include observations made in different situations or by different people at different times or data from different assessment instruments. However, data from more than one of the same kind of assessment instrument (for example, a series of standardized tests) will not satisfy this standard because such tests commonly reflect a similar and narrow view of literacy. By the same token, even new data can be looked at with old eyes. Unless different perspectives and values are brought to bear on data, our understanding might not expand. Even the richest set of data can be reduced to mere conventions by a limited perspective.
From a more statistical point of view, the reliability of interpretations of assessment data is likely to improve when there are multiple opportunities to observe reading and writing. Adherence to this standard will also substantially improve the validity of the literacy assessment process because sampling more than one aspect of literacy permits a closer approximation of the complexity of reading, writing, listening, and speaking processes as they occur and as they are used in real-life settings.
However, seeking multiple perspectives and sources of data is not intended only for the purposes of reducing biases or errors in individual data sources. Instead, it takes advantage of the depth of understanding that varied assessment perspectives afford and the dialogue and learning they produce. Two teachers with different cultural or linguistic backgrounds might interpret a student’s literacy development in different ways, each of which provides an important perspective. Indeed, because literacy learning is also social in nature, these two teachers’ different interpretations will lead to different kinds of development. The exploration of these contrasting perspectives will lead not only to a more productive understanding of the specific student’s development but also to an enhanced awareness of possible interpretations of other students’ development—and of what it means to develop.