had at least seven measurements across the school year), a preliminary analysis assessed the extent to which a linear relationship adequately modeled student progress within 1 academic year of development. This analysis was conducted on Years 1 and 2 reading, to describe the oral passage reading and the maze reading measures, and on Year 2 spelling and math data. As slope was calculated for each datum for each of these individuals at every grade level, a quadratic component was included in the analysis to determine whether it contributed to the modeling of student progress. The percentage of students for whom the quadratic term contributed significantly was computed for each datum at every grade level. The percentage of students for whom the linear term significantly contributed also was computed.
Examination of distribution of slopes. After we calculated the individual slopes and estimated the extent to which quadratic and linear relationships modeled student progress, the mean slope and standard deviation were calculated for each datum at every grade. Then, the distribution of the slopes by grade and datum was examined to assess conformity to the normal distribution. Kurtosis and skewness were calculated, and the percentage of distributions for which kurtosis or skewness exceeded two standard errors was computed. The percentage of cases with negative slopes also was calculated.
Analysis of the relationship between slope and grade level. Finally, the effect of grade level on slope was examined by running a one-way analysis of variance (ANOVA) for every datum. The ANOVA employed grade level as the between-subjects factor and incorporated orthogonal polynomial contrasts to determine the extent to which a linear or quadratic relationship modeled student progress across years of development.
In the summer preceding Year 1, initial training was delivered for 2 days to the districts' special education coordinator, who supervised the project, and to three contact teachers (i.e., one general educator in each participating school). During the first day, a rationale for the project was presented, the measurement procedures were explained and practiced, and methods for interpreting CBM graphs were reviewed. On the second day, measurement and graph reading were practiced again, questions were answered, and each participant was provided scripted materials and procedures with which to train the remaining general educators in her home school.
In the second training phase, which occurred in September of each study year, the scripted training session was conducted at each school by the special education coordinator and that school's contact teacher. This training session included (a) a rationale for the measurement, (b) instruction and practice in administering CBM tests, and (c) instruction and practice in CBM graph reading.
In the third phase, the special education coordinator held monthly inservices each year, at which concerns were discussed, measurement procedures reviewed, and student graphs shared and discussed. Additionally, the contact teacher at each school provided ongoing support and served as the consultant for the project at her school.