Information Technology Reference
In-Depth Information
Fig. 10.1
Connections between research, test design, and practice. Adapted from National
Research Council (2001, p. 295)
learning process. Thus, tests not carefully wedded to a theory of learning can be
used to classify students and schools, to sort them, but such tests cannot be used
as educative aids, because they provide no insight into the learning process per se.
On the other hand, building tests around an explicit theory of learning increases the
range of functions the test can serve, for example, allowing for insight into what a
student has learned and could most benefit from learning next.
Contemporary theories of learning have become increasingly sophisticated due to
advances in cognitive science and neuroscience. Likewise, advances in psychomet-
rics have made it possible to reliably and validly measure a wider range of dynamic
and meaningful constructs. As alluded to in the historical section above, the contem-
porary testing infrastructure, and the uses to which it is put, reflect these advances
only in a very limited way, if at all. The rest of this chapter is devoted to outlin-
ing one approach that pulls together advances in the new science of learning and
psychometrics to retool test design and educational practice.
Fischer's
Dynamic Skill Theory
(Fischer & Bidell, 2006) provides a compre-
hensive and empirically grounded approach to human development and learning.
This approach has fostered methods for building empirically grounded learning
sequences in a variety of domains, which can be aligned along a common underlying
developmental dimension. This underlying dimension has been operationalized as
a psychometrically refined metric, known as
The Lectical Assessment System
(the
LAS: Dawson, 2008), which has been used to build unique and richly educative
tests that are both standardized and customizable to different curricular frameworks
(Dawson & Stein, 2008). The
Discotest Initiative
is the name given to our efforts in
this direction. Below we discuss this approach to redesigning standardized testing
infrastructures based on the new sciences of learning; we explain why it should be
seen as a valuable alternative to the infrastructures currently in place.
However, the value of different testing infrastructures should not be determined
solely on the basis of the manner in which they wed research about learning with
test design and educational practices. Decisions must be made about the general
shape of the educational system and the role that ought to be played by even the
best-designed tests. These decisions are fundamentally evaluative. They touch on
some of our broadest goals and commitments about how schooling fits into a shared
vision of the good life and just society. Typically, a set of classic dichotomies have
been used to frame this discussion: Should tests be used to sort individuals and make
high-stakes decisions or should tests be embedded in the curricula as educative aids?
Should tests assess broad competencies or specific knowledge? Should tests help us
in training the workforce or in fostering critically minded citizens?
