The data sgp package contains classes, functions and data used to conduct student growth percentile (SGP) analyses. SGP analyses utilize large scale, longitudinal education assessment data to determine percentile growth projections and trajectories for individual students based on their performance history. This is accomplished by applying a statistical procedure called quantile regression to place the individual’s performance on a normative scale and showing the amount of growth necessary for the student to reach future achievement targets.
The SGP package makes it easy to run these analyses. However, like all analyses the bulk of the time spent conducting SGP analysis is in data preparation. The data sgp package has several higher level wrapper functions (abcSGP, prepareSGP, analyzeSGP) that “wrap” the lower level SGP function studentGrowthPercentiles and studentGrowthProjections simplifying the source code associated with running these analyses. For operational SGP analyses it is usually more efficient to use the wrapper functions rather than the lower level functions when performing a series of SGP analyses in an operational setting.
SGP trajectories are a valuable educational tool for teachers, students, parents, and districts. However, the process of calculating these projections can be complicated and time consuming. For this reason OSPI has developed a set of tools to simplify the process of generating SGP projections and trajectories for student groups. These tools are available for download on the Student Growth District and School Resources webpage.
To access the SGP projections and trajectories you need to have a computer with the free, open source software environment R installed. R is available for Windows, OSX, and Linux and is freely downloadable from the CRAN website. There are a number of resources for getting started with R including a quickstart guide and tutorials on the R website.
In the Star Growth Report window specific SGPs are calculated for each student based on their performance on up to two previous spring MCAS tests. The student’s SGP score is compared to that of their academic peers who scored similarly on the 2024 and 2023 exams using a statistical procedure called quantile regression. This allows us to place the student’s score on a normed scale and identify how much growth is required for the student to achieve proficient status in their particular subject area.
Average SGP scores are calculated statewide and for schools, districts, and various student subgroups such as gender, race, income, and educational programs (e.g., sheltered English immersion, special education). These averages can fluctuate slightly because they are based on a smaller sample of available data. For example, the average mSGP score of an English language arts teacher in grade 6 during the 2024 spring will be slightly different than the average mSGP score of the same teacher during the 2022 spring. This is because the 2024 cohort is a larger sample than the 2022 cohort. The average mSGP scores for each cohort will be adjusted to compensate for this difference.
