The data sgp package contains functions to calculate student growth percentiles and projections/trajectories from longitudinal assessment data. These functions use long formatted data, where each case/row represents a single student and each column is a time dependent variable. Data sgp provides the exemplar WIDE and LONG formatted data sets sgptData_WIDE and sgptData_LONG to assist users with setting up their own longitudinal (time dependent) assessment data for use in SGP analyses.
SGPs compare students’ performance against that of academically-similar peers in the United States. SGPs can help educators assess a student’s progress toward meeting achievement targets or other goals set by teachers and administrators.
In addition to SGPs, the data sgp package includes tools for creating a number of other student-level assessments and reports. These tools allow educators to measure a student’s current performance, monitor their progress over time, and identify areas for improvement.
For example, teachers can measure a student’s current progress on a state assessment by comparing the student’s score on their most recent Star assessment to the same student’s score on the student’s previous test in another testing window: Fall to Winter or Spring to Summer. Educators can also create an individual learning plan for each student based on the student’s score on the most recent Star assessment. This plan can include recommendations on how the student can improve their score and a timeline for implementing these recommended strategies.
The data sgp package also includes a set of tools to visualize and explore growth trends over time, including a graph tool, a trend line for each student’s score on the most recent assessment, and an interactive map. This allows educators and researchers to visually inspect and interpret growth patterns for a student’s score on the most recent assessment as well as trends over time for their entire cohort of students.
Another tool available to educators is the ability to generate Goodness of Fit figures for an SGP model. These figures can be downloaded as PDF files and can be used to evaluate the model for accuracy.
Lastly, the data sgp package contains a number of additional functions to perform state-specific analytics such as accessing embedded knot and boundaries, cutting scores, CSEMs, and other state related assessment data. The data sgp package is also designed to be extensible, with new functions being added as they are required by educators.
A good rule of thumb is to have 4GB of memory available per core for running large state data (e.g., sgp.percentiles and sgp.projections).
The sgp.short() function returns an object of class SGP which contains long formatted data in the @Data slot (from prepareSGP).