All of the research recommended by the RRSG depends on having better in-
struments for assessing reading comprehension. The impact of assessment on
instruction constitutes a research agenda of its own, particularly in the current
era of accountability-oriented education reform. A system of reading compre-
hension assessment should reflect the full array of important comprehension
outcomes and a research program should establish appropriate levels of perfor-
mance for children of different ages and grades based on those outcomes. With-
out research-based benchmarks defining adequate progress in comprehension,
we as a society risk aiming far too low in our expectations for student learning.

The RRSG proposes an approach to assessment that differs from current ap-
proaches in that it is based on an appropriately rich and elaborated theory of
reading comprehension. The assessment procedures in this approach will be
fluid, and they will change as more is learned from the research. More value will
be placed on their usefulness for improving instruction. And because compre-
hensive assessment systems can place significant time demands on students
and teachers, the education community has an obligation to develop assess-
ments that are an integral part of and supportive of instruction, rather than lim-
ited to serving the needs of researchers.

Teachers who are interested in improving their instruction need reliable and
valid assessments that are closely tied to their curricula so that they can identify
those students who are learning and those who need extra help. The compre-
hension assessments that are widely used today focus heavily on only a few
tasks and thus may inadvertently limit the reading curriculum to preparation
for those few tasks. Knowledge, application, and engagement are all critical
outcomes of reading with comprehension; assessments that reflect all three of
these outcomes are needed.

-xix-