To illustrate how this type of normative information can be used within special education settings, we briefly describe a case study about a sixth-grade student who is receiving instruction in a special education resource room and is working on the fifth-grade math curriculum this year. The CBM system is designed to mirror this curriculum: Each 25-item parallel test requires the student to complete every type of problem to be taught during the year. The student takes two parallel forms of the test each week. The primary CBM summary of performance is a graph showing the student's total test scores (i.e., number of digits correct) over time (see Figure 1).
On his first three CBM tests sampling the fifth-grade mathematics curriculum, the student earned CBM scores of 37, 29, and 24 digits correct. Using a baseline 29 digits correct (i.e., the median of his first three scores), the school psychologist could consult norms for weekly rates of growth to determine that, based on a general education sample including 21% handicapped pupils, fifth graders typically improve approximately. 75 digits per week (see Table 6). The school psychologist realizes, however, that this special education student must improve upon the typical rate of progress in order to reduce the discrepancy between his performance and that of his peers. Consequently, instead of using the more typical growth standard of .75 digits improvement per week, the school psychologist employs the more ambitious criterion of 1.2 digits per week (i.e.,.75 plus one standard deviation, .45). Because 26 weeks remain in the academic year, the year-end goal then is set at 61 ([1.2 x 26 weeks] + a baseline of 29). This weekly rate of improvement is depicted on the student's graph with a broken diagonal line.
The teacher and school psychologist can use this CBM structure to evaluate the student's progress in the 95th-grade curriculum and to plan his instructional program. When the student's rate of growth exceeds 1.2 digits per week, the goal is increased. As shown in Figure 1, however, the student's rate of growth is less the desired criterion of 1.2 digits per week: The solid line superimposed over the last 10 scores indicates that the actual rate of growth is less than the desired rate, reflected in the broken diagonal line. Consequently, the teacher adjusts the teaching program to try to effect better growth (see message at the bottom of the graph suggesting a teaching change). In consultation with the school psychologist, the teacher relies on information about the instruction provided, about the student, and about the CBM database to determine how to adjust the student's program.
The CBM database also can be used to account for student outcomes. Between November and May (26 weeks), the student progressed from 29 to 75 digits correct on the Grade 5 curriculum, representing a weekly increase of over 1.5 digits. This slope of 1.5 digits compares favorably to the student's prespecial education slope of .75 for the mainstream 95th-grade class working in the Grade 5 curriculum.
In general education CBM implementations (e.g., Fuchs, 1992), every student in the class is measured once each week on the grade-appropriate curriculum. This general education CBM database can be used to identify students who are experiencing difficulty with the curriculum early in the year, when their weekly rates of growth fall below normative targets for growth. These students can be identified for special attention within general education. Sometimes,