Data SGP is a software package developed for the R statistical software environment. It provides classes and functions for performing student growth percentiles (SGPs) and percentile growth projections/trajectories using large scale, longitudinal education assessment data. SGPs provide a view of how students are progressing over time compared to their academic peers and are a powerful tool for identifying high achieving students as well as those that may require additional support. It also allows educators to articulate goals and achievement targets more effectively by describing how much student growth is necessary to achieve the target.
Unlike other methods for assessing student growth, data SGP is a longitudinal measure of student achievement that compares the performance of students on statewide tests to their academic peers. It uses a combination of the most recent Star assessment and one prior test from an earlier testing window to determine how students are progressing toward the district or school goal for proficiency. In addition, SGPs provide educators with a clear picture of the growth trajectory that students must take in order to reach their proficiency target and other goals they set for themselves.
SGP analyses are designed to be as simple and straightforward as possible, but they are a two step process – the bulk of the work occurs in the preparation of the data for analysis. If the data is not prepared correctly, even a small number of errors could lead to incorrect or misleading results. To help simplify the preparation process, the following steps are outlined in this article.
The first step in preparing data for SGP analysis is to identify all of the assessments that are valid for the current year. To do this, districts must review the course roster submission that is submitted through NJ SMART in the summer following each year’s statewide assessments. The roster lists the students in each class and indicates the teacher of record for each student. The teacher of record must be the teacher that taught the majority of the course in order for a student to be eligible to receive an mSGP score.
Data SGP analyses are only meaningful if they are based on the most current and complete assessment data available. To ensure that this is the case, a database must be built and maintained in accordance with the requirements of the data SGP functions. This requires a robust data governance process that includes creating and maintaining the required tables, ensuring data quality and establishing auditable processes to verify that the required data has been analyzed.
The database used to perform SGP analyses must be formatted in the WIDE or LONG formats depending on the analyses being conducted. To illustrate how to prepare and use this data, an example SGP data set has been created and is provided below. The sgptData_WIDE dataset models the WIDE data format that is used by lower level SGP functions such as studentGrowthPercentiles and studentGrowthProjections. The sgptData_LONG dataset has been designed to be used with higher level wrapper functions that leverage the LONG data format which offers numerous preperation and storage benefits.