Information Technology Reference
In-Depth Information
officials. These recent developments have increasingly brought testing into the cen-
ter of national debates about education. In particular, the Obama administration has
drawn attention to the liabilities of the contemporary testing infrastructure—in both
its specific details and broad impacts (Obama, 2008; White House, 2009).
Teaching to the tests as they currently exits means preparing students for life as
if it were a set of multiple-choice questions. It would seem that the time is ripe for
seriously questioning the foundations of our testing infrastructure, asking a series
of very basic questions:
What is being measured with today's tests? What should be
measured? How are today's tests being used? How should they be used?
Reflecting on Testing: The Need for a Theory of Learning
and Clarity About Values
Although it may not be apparent at first, questions about what we are measuring
with tests and how they should be used hinge upon the way we conceive the nature
of
learning
(NRC, 2001). The criticism of contemporary tests as the application of
advanced statistics in service of simplistic psychology is to the point. In order to
use tests effectively and knowledgeably, we need to understand the meaning of the
score a student receives. Does a score mean something about a capability or trait
processed by the student, or does it simply let us know how the student performs
on a specific set of questions in relation to group averages? The latter claim—given
that it remains strictly descriptive, positing no explanation for the score (e.g., IQ)
or prescription for changing it—does not entail beliefs about specific psychological
constructs; the former does.
Claiming that a test score reflects an underlying capability or trait—be it an
aptitude, skill, or disposition—entails certain views about these psychological
phenomena. More specifically, using such a claim about the meaning of a test
score to guide actions, such as doling out remediation or rewards, entails some
theory of
learning
. Imagine again the high school student from Massachusetts tak-
ing a standardized test. The use to which her scores will be put—determining
graduation eligibility and school standing or quality—imply that the capabilities
being assessed are the result of her individual effort and the school environment.
That is, they assume a theory—however implicit—about how the capability being
measured changes over time as a result of certain factors. Roughly speaking, theo-
ries about how psychological phenomena undergo change are theories of learning or
development (Reisberg, 2001). Different theories of learning will give the same test
score different meanings, and different theories of learning result in different forms
of test design (NRC, 2001; see Fig. 10.1). Moreover, a test built without an explicit
theory of learning—as many tests are—can serve only very limited functions.
For example, most standardized tests, like the one taken by the Massachusetts
student, can serve
only
as sorting mechanisms because they are built without
reference to an explicit theory of learning. No doubt, they are reliably and objec-
tively measuring
something
, but it is unclear how this “something” relates to the
