CBM Maze measures provide a practical tool for monitoring reading progress because they can be administered at a group level and because they can be administered and scored with a relatively inexpensive computer program, Monitoring Basic Skills Progress, available from Pro-Ed.

Research has been conducted to examine the technical adequacy of CBM as a progress monitoring system; however, most research as been based on static measures obtained at a single point in time, not on growth measures of student performance.

This study was designed to extend the recent research on the technical adequacy of repeated CBM for quantifying growth. Three specific psychometric characteristics immediately and practically important for scaling growth were examined in this study: (a) alternate-form reliability, (b) sensitivity to actual changes in the academic skills being developed, and (c) criterion validity of the growth rates estimated through repeated CBM.

Forty-three second graders participated in the study.

"Results provide evidence that the maze task reliably assesses student growth in reading based on repeated performance measures collected at intervals from 1 to 3 months over an academic year." p. 168

Maze can detect significant group improvement as well as changes in an individual student's performance, even for students in the lower elementary grades.

Maze is predictive of performance on other important measures of reading achievement.

"…low maze growth rates could be viewed as a concern for progress monitoring. The magnitudes of maze growth rates are low (e.g., 1.07 increases per month for second graders in this study) relative to those obtained through reading aloud (e.g., 9.14 increases per month for second graders converted from the weekly growth rate reported by L.S. Fuchs at al., 1993). As a result, the maze task might not appear to be as sensitive in assessing growth over time as reading aloud. L.S. Fuchs and Fuchs (1992), however, found that there was compatibility in growth rate estimates between the maze and reading aloud tasks, and that maze growth rates could be converted to reading aloud growth rates." P. 170

Researchers found evidence of "the Matthew Effect" (the good readers getting better and the poor readers getting worse) in this study. They noted that variance of maze scores increased over time and that correlation between initial status and growth rates was very high (.97).

"...further research on student growth with multiwave maze measures is needed to (a) identify the growth patterns for students receiving general and special education services (Shin, 1999), (b) estimate typical growth rates for these two populations at different grades (L. S. Fuchs et al., 1993; Shin, 1999), and (c) establish functional relations between growth rates and instructional or individual characteristic variables." p. 171

Questions or comments can be sent to: Sue Dungan

Return to Heartland Home Page

© Heartland AEA 11, 2001