Building block 4: Wright maps. Wright maps represent the principle of highquality
evidence. Progress maps are graphical and empirical representations of
a progress variable, showing how it unfolds or evolves in terms of increasingly
sophisticated student performances. They are derived from empirical analyses
of student data on sets of assessment tasks. Maps are based on an ordering of
these assessment tasks from relatively easy tasks to more difficult and complex
ones. A key feature of these maps is that both students and tasks can be located
on the same scale, giving student proficiency the possibility of substantive
interpretation, in terms of what the student knows and can do and where the
student is having difficulty. The maps can be used to interpret the progress
of one particular student, or the pattern of achievement of groups of students,
ranging from classes to nations.
Wright maps can be very useful in large-scale assessments, providing information
that is not readily available through numerical score averages and other
traditional summary information—they are used extensively, for example, in
reporting on the PISA assessments [PISA 2005a]. A Wright map illustrating
the estimates for the Rasch model is shown in Figure 4. On this map, an “X”
represents a group of students, all at the same estimated achievement level. The
logits (on the left-hand side) are the units of the Wright map—they are related
to the probability of a student succeeding at an item, and are specifically the log
of the odds of that occurring. The symbols “T,” “N” and “A” each represent a
Technical Processing, Numerical Modeling, and Abstract Modeling item, with
the Topic Area indicated by the column headings above. Where a student is
located near an item, this indicates that there is approximately a 50% chance of
the student getting the item correct. Where the student is above the item, the chance is greater than 50%, and the further it is above, the greater the chance.
Where the student is lower than the item, the chance is less than 50%, and the
further it is below, the lesser the chance. Thus this map illustrates the description
of the Mathematical Literacy progress variable in terms of the Levels from page
316 as well as the Topic Areas and the Modeling Types in the items design. The
Topic Areas reflect the earlier placement of Arithmetic in the curriculum than
Geometry and Algebra. The ordering of Modeling Types is generally consistent
with what one might expect from the definitions of the Levels, except for the
Arithmetic Abstract Modeling items, which seem to be somewhat easier than
expected. This is a topic that deserves a follow-up investigation.
We typically use a multi-dimensional Rasch modeling approach to calibrate
the maps for use in the BEAR Assessment System (see [Adams et al. 1997]
for the specifics of this model). These maps have at least two advantages over
the traditional method of reporting student performance as total scores or percentages:
First, it allows teachers to interpret a student’s proficiency in terms
of average or typical performance on representative assessment activities; and
second, it takes into consideration the relative difficulties of the tasks involved
in assessing student proficiency.
Once constructed, maps can be used to record and track student progress and
to illustrate the skills that students have mastered and those that the students
are working on. By placing students’ performance on the continuum defined by
the map, teachers, administrators, and the public can interpret student progress
with respect to the standards that are inherent in the progress variables. Wright
maps can come in many forms, and have many uses in classroom and other
educational contexts. In order to make the maps flexible and convenient enough
for use by teachers and administrators, we have also developed software for
teachers to use to generate the maps. This software, which we call GradeMap
[Kennedy et al. 2005], allows consumers to enter the scores given to the students
on assessments, and then map the performance of groups of students, either at
a particular time or over a period of time.