Date of this Version
After closely examining the recent history of reading comprehension assessment in the United States, we have concluded that although both the forms of assessment and the key players in the assessment process have changed in significant ways, the functions of assessment have remained relatively constant. In terms of function, we have always used, and continue to use, assessment tools to evaluate programs, to hold particular groups accountable for some specified set of outcomes (though it may seem that that is all we do these days), to inform instruction, either for individuals or whole classes, and finally, to determine who gains access to particular programs or privileges (the gatekeeping function). However, very different test formats, or at least a very different mix of formats, are used today than were used twenty-five years ago. We contend that changes in our fundamental views of the reading process have paved the way for these new formats. We argue that changing and sometimes conflicting policy contexts (what legislators and other policymakers want from assessments) have been responsible for shifting an emphasis from some functions (e.g., instructional decision making) to others (e.g., accountability) and have changed who it is that decides who shall take what tests and for what purposes.
We also attempt to document another thesis, one that is more interpretive than descriptive: Progress, if one can even characterize the history of reading comprehension assessment as moving in a particular reform-minded direction, is best characterized as "two steps forward, three steps back." Usually, a forward step is an advance in assessment practice driven by an advance in reading theory, or possibly psychometric theory. Usually, the backward step is a retreat in assessment practice driven by some political or practical constraint. As we discuss later, the most notable retreat in the last quarter-century has been in the area of accountability. In the name of holding schools and teachers responsible for student performance, education officials have created such a high-stakes environment that people end up "teaching to the test" in a way that narrows rather than expands curricular opportunities. A second "step back" has been the retreat in the use of portfolios and performance assessments; they are considered either too personal (a political motive) or too time-consuming for the quality of information obtained (a practical motive).
We make these two points by examining the historical course of reading comprehension assessment practices over the last quarter-century. To understand the current mix of comprehension assessment practices, we believe that it is necessary to begin with a characterization of the assessment practices that were dominant in the 1970s and then to work our way to the present, trying to understand each new assessment twist in light of changing views of reading processes, practices, and policies.