What is Data SGP?

data sgp

Data SGP is a tool designed to analyze longitudinal student assessment data and produce statistical growth plots (SGPs), providing visual evidence of student progress relative to academic peers. SGP analyses utilize latent achievement trait models and compare students’ current assessment scores against set growth standards established from their past testing history – more accurate measures of student growth than traditional percentiles provide.

SGPs are calculated through a mathematical process known as quantile regression that estimates the conditional density associated with each student’s score history, drawn from their covariance matrix of previous raw and scaled scores. Coefficient matrices created during this process are then used to compute students’ growth percentiles as well as projections or trajectories for growth percentiles.

Students in Wyoming are compared with their academic peers – those students with the same test-score path that have taken similar tests over time (for instance, taking multiple sections at one time). SGPs allow educators to gauge how their students compare to others of similar age in Wyoming and identify areas of support needed to ensure student achievement.

SGPs are increasingly being used by high schools across the nation as evaluation metrics, yet educators must keep several things in mind before beginning to interpret this new metric. While SGPs measure progress, they cannot tell us whether a student’s growth is appropriate nor whether their current performance meets standards.

Calculating SGPs requires complex calculations; however, the information can be shared using familiar percentiles to allow teachers and parents to quickly identify areas in which a student may be struggling or extra support may be required. This is essential in providing schools with information for targeting areas that need special assistance for individual students.

The SGPdata package offers an example WIDE format data set called sgpData_LONG that can be used to simulate time dependent data for lower level functions like studentGrowthPercentiles and studentGrowthProjections. Higher level functions wrappers for these functions are included within SGPdata to simplify operational SGP analyses’ source code.

Educators should also be mindful that the Student Growth Plans released by MDE are only preliminary, and may not yet be suitable for educator evaluations in 2015/16 and any high stakes decisions until 2018. Therefore, it is strongly suggested that SGPs NOT be used for educator evaluations until 2018.

For any inquiries related to SGPs, we invite you to reach out to either the Wyoming DEPD’s Professional Development Center at PDCenter@wyoming.edu or its SGP & Equity Team at SGP@WYOmING.edu – they’d be more than happy to help get you going with these powerful new tools! We hope we see you soon.