What is Student Growth Percentile (SGP)?
A student’s growth percentile (SGP) describes a student’s performance relative to their academic peers. SGPs are calculated using large scale, longitudinal education assessment data. This data contains information on the student’s prior test scores, their current achievement levels and their predicted future achievement levels based on past performance patterns. SGPs are important because they provide valuable information about a student’s progress at all performance levels. This information is communicated in terms that are familiar to teachers and parents.
Students who are struggling can be identified and supported with targeted instruction and supports. Students who are performing well can be encouraged to reach even higher achievement levels. Growth measures help all students, including those at the bottom of the curve, to show progress. Typical SGPs are between about 36 and 64, so students or groups with greater or lower SGPs may not be educationally meaningful.
SGPs can be used to compare student outcomes and school and district performance across time and across districts. By providing a common framework for measuring student achievement and growth, SGPs allow comparisons of student progress with the achievement of similar students in different regions of the country. This allows schools to better evaluate their programs and make strategic decisions about how best to serve all students.
To use SGP analyses, the user must have access to the data and the appropriate software. This is why we are excited to announce the release of the data sgp package that makes it possible to quickly and easily run SGP analyses from within Python. This package is available as a standard distribution of the statistical programming language, and it provides an easy way for users to set up their own data sets and conduct SGP analysis without writing code or downloading other software packages.
Data sgp is a collection of classes and functions that perform Student Growth Percentile (SGP) analyses. The package also includes a series of exemplar data sets for demonstrating how to set up and use the SGP package. The data sets sgpData, sgpData_LONG, sgptData_LONG and sgpData_INSTRUCTOR_NUMBER are WIDE formatted data sets that can be used with the lower level functions studentGrowthPercentiles and studentGrowthProjections. However, for operational analyses it is recommended that the LONG format be used as this provides many preparation and storage benefits over WIDE.
As long as the proper steps are taken in preparing the data, SGP analyses are usually straightforward and error free. Any errors that are found in the data usually revert back to issues with data preparation. It is always worth checking the data carefully for potential errors and ensuring that it is set up in the right format before proceeding with an SGP analysis. This will save you a lot of frustration and time.