What Is Data SGP?

Data sgp is a student growth percentile analysis software tool that can be used by school districts, teacher unions, and other stakeholders to access and analyze growth information provided by the Massachusetts Department of Education (DEE). SGPs measure students’ progress on statewide assessments relative to their academic peers.

SGPs can be calculated for both current and prior assessment windows. SGPs for students with one or more prior tests are based on the average of the students’ performance on those previous tests and their SGP for the current assessment window. This SGP for the current assessment is then compared with their previous SGP to identify which student’s performance is higher or lower than the others’.

A student’s SGP is a single number that indicates the extent to which a students’ current score exceeds or falls short of the 75th percentile of their academic peers. SGP scores are reported on a scale of 1 to 99, where a higher number indicates greater relative growth.

While a student’s SGP may fluctuate from year to year, the underlying model is designed to be stable over time. This is because the underlying model is a statistical regression model with an error term that is a function of the difference between the test score and the 95th percentile for that particular test. This means that the resulting change is the expected value of the student’s growth, which can be used to predict future test results for each individual student.

SGP analyses can be conducted in two ways, with the bulk of the work involved in conducting these calculations being in data preparation. The SGP package provides wrapper functions, abcSGP and prepareSGP, that simplify the source code needed to conduct these analyses. The examples contained in this article use these two wrapper functions to illustrate SGP analyses.

To run SGP analyses, the user will need a computer running the open source statistical software program R. This software is available for both Windows, OSX and Linux and can be downloaded from the CRAN website. In addition to R, the SGP package requires a large amount of data that is stored in a specific format. This data can be accessed by the SGP package through its SGPdata data set.

The SGPdata data set contains 4 examplar data sets for use with SGP analyses. The first, sgpData_WIDE, specifies the data in the WIDE format required for the lower level SGP functions studentGrowthPercentiles and studentGrowthProjections. The other three datasets in the SGPdata package, sgptData_LONG, sgpData_INSTRUCTOR_NUMBER and sgpData_teacher_lookup, provide the data in the LONG format that is required by higher level SGP functions like prepareSGP and analyzeSGP.

The SGPdata data set includes exemplar SGP analyses that can be used to explore the underlying processes that are involved in SGP calculations. Using these examples, users can better understand the SGP methodologies and learn how to conduct their own SGP analyses. The example analyses include explanations of each step of the process, as well as the underlying data set and output.