Exploring the generalizability of curriculum-based measurement progress monitoring procedures in reading across students in general and special education

The purpose of this study is to examine educational decision-making utilizing Curriculum-Based Measurement (CBM) through a Generalizability theory (G theory) model. Prior research (Martson, 1989, Shinn, 1989) clearly demonstrates that CBM procedures are reliable and valid. However, the classic test score theory, used in the research to determine reliability, describes the measurement error as random and without any specific context. G theory extends classic test score theory by attributing the error to various measurement components. Previous research (Hintze & Pelle Petitte, in press; Pelle, Hintze, & Owen, 1998) found that approximately 66% of the variance was attributed to individual differences, while group differences accounted for only 15% of the variance. Furthermore, this effect was stable over time. ^ Sixty-seven students (in grades 2, 3, and 4, from a school located in the Northeast, U.S.) read orally for one minute each over sixteen trials. Thirty-four of the students were receiving or had been receiving general education reading instruction only, while the other thirty-three were receiving or had received some form of remedial reading instruction. The increased sample size used in this study adds to the probability of determining statistical significance of the empirical findings. ^ The results from this study are consistent with prior research (Hintze & Pelle Petitte, in press; Pelle et al., 1998), adding to the robustness of CBM and G theory literature. Conclusions from the analysis demonstrate that CBM procedures are generalizable across remedial and non-remedial groups of students. In addition, CBM procedures are also shown to account for adequate variability between the two groups of students and across time. These findings support the wider of CBM as a means of evaluating students based on their current curriculum as opposed to annual standardized assessment.