Data sgp is an analysis tool for the statistical software environment R that allows users to create statistical growth plots, or SGPs. These plots provide information about students’ relative progress compared to academic peers and can be used as a benchmark against which a student’s performance can be measured. SGPs are calculated from a student’s longitudinal (time dependent) standardized test score history.
A key part of the SGP is that it shows how much a student needs to grow in order to meet an achievement target or remain proficient, and thus serve as a realistic measure of progress for each individual. SGPs can be used to support educators in articulating and measuring goals for students and in evaluating the effectiveness of their programs. Moreover, SGPs can be used to identify high achieving students as well as those who need additional support to meet their academic goals.
To develop an SGP, DESE compares a student’s current test score with a number of previous scores from their grade-level tests in ELA and math. For example, a student scoring 300 on this year’s test would be compared with their prior test scores in grades 3 through 8. SGPs are not produced for science because of the lack of historical data for that subject.
When a student’s SGP is determined, the student is assigned a percentile rank. This number indicates what percentage of academic peers scored higher on the test. For example, a student scoring 300 this year would be ranked 75th out of their academic peers who scored the same on the test. The higher the percentile rank, the more a student has achieved relative to their academic peers.
The SGPdata package includes the SGP analyses functions prepareSGP, analyzeSGP, and combineSGP that allow users to conduct operational SGP analyses in a simple and straightforward manner. These functions require exemplar longitudinal data sets in WIDE and LONG format, and the Demonstration_SGP@Data data set is installed with the SGPdata package for this purpose.
Using these functions, a user can prepare a data set for analysis by creating a master longitudinal record for each student, including an indicator for whether they are a student of color or have disabilities. This record is then analyzed for the SGP targets and projections desired by the user. Once the results of these analyses are obtained, they can be combined to produce an SGP report for a specific window by year by content area and grade grouping.
Most of the time that users spend conducting SGP analyses is spent on data preparation. In our experience, nearly all errors that come up during data analysis revert back to issues in the data preparation process. For these reasons, we try to keep the data preparation and analysis steps as simple and straight forward as possible. We also have a number of online resources available to assist newcomers to SGP analysis. A good starting point is the SGP Tutorial. If you have further questions, feel free to contact us or post an issue on GitHub.