The data sgp software is an analysis tool for the statistical software environment R. It is free to use and includes numerous online resources that assist newcomers to the program with data preparation and analysis. It is primarily used with educational assessment data such as MCAS test scores but can be applied to any type of statistical data set.
Data sgp provides an alternative to calculating standardized test score growth percentiles from raw student data by combining students’ prior and current MCAS performance into a single, more accurate measure of relative student achievement. This method is also more meaningful than comparing students’ raw scores since the latter do not account for changes in a student’s ability due to learning or other factors.
The data sgp program performs several steps to prepare, analyze and project student growth percentiles and projections. The process requires a significant amount of data preprocessing. Most errors encountered during the analysis process typically revert back to problems with the data preparation process. Therefore, it is important that districts devote ample time to properly preparing data for analysis.
First, students’ prior test score records must be combined into a common scaled record for each student. Then, the student’s current test score is compared with this scaled record to calculate the student’s progress in relation to their peers. This information is then projected into the future to allow for comparisons between different cohorts and years of testing.
Using the MCAS student data, DESE produces growth percentiles for ELA and math for students in grades 4 through 8, and for science in grade 10. The growth percentiles are based on students’ current test scores being ranked against the results of students with similar MCAS performance histories (i.e., their prior test scores).
The data sgp software is designed to facilitate these processes for districts in the Macomb and Clare-Gladwin ISDs. The software package is distributed as a git repository and is available for download from the RStudio website. The package’s documentation, vignettes and examples provide detailed explanations of the calculations and processes involved in SGP analyses. Additionally, the higher level wrapper functions abcSGP and updateSGP simplify the source code associated with operational SGP analyses by combining the six lower level functions into a single function call. In addition, the sgpData_INSTRUCTOR_NUMBER lookup table is an invaluable feature that allows districts to connect students with instructors through unique identifiers in their test records. This facilitates the efficient processing of SGP analyses and allows for the comparison of students across instructors, schools, and grades. The results from this data sgp program are powerful tools that can help teachers, administrators, and policymakers identify students who need additional support. This in turn can help inform policies and practices that will ultimately lead to increased student achievement. The true value of these efforts, however, is difficult to evaluate unless a district’s data preparation process is optimized. If not, these valuable insights can be misleading.