Data sgp is a software package for analyzing longitudinal student assessment data to produce statistical growth plots (SGP). SGPs display students’ relative progress in comparison to their academic peers. Educators use SGP data to identify students that require additional support, differentiate instruction for high performing students and monitor whether students are on track to meet achievement goals.
The classes and functions in the SGP package leverage large scale, longitudinal education data to create student growth percentiles and associated percentile growth projections/trajectories using quantile regression. This method is more accurate than other standard methods for SGP analysis such as mean, median and mode score calculations. SGPs can also be used to assess student progress toward a target score, for example a 75% of their academic peers in a given content area.
To run SGP analyses, a user must have the R software environment installed on their computer. R is a free and open source programming language that runs on Windows, OSX or Linux. While the SGP package contains some advanced functions, running SGP analyses is generally straight forward. Most errors that arise when attempting to run SGP analyses revert back to issues with the initial preparation of the data set.
When preparing SGP data for analysis, educators must ensure that their student assessment data is correctly formatted. This is necessary because the lower level SGP analysis functions, such as studentGrowthPercentiles and studentGrowthProjections, require WIDE formatted data. The SGP package also provides higher level wrapper functions that are designed to work with LONG formatted student assessment data sets.
LONG formatted student assessment data is typically stored in a tabular format where each row represents a unique student and each column represent a particular content area assessed over time. The example sgpData file above, which is included in the SGP package, has 5 years of student assessment data and includes columns for student ID, grade level, subject area and teacher (instructor) identification.
In general, when working with long data sets, one should always consider the possibility of missing values – that is, students who did not take an assessment in a particular year and thus cannot be compared to their peers. Missing values should be appropriately flagged and handled when constructing student growth percentiles and projections.
SGP analysis is designed to be as simple as possible for users. However, running these calculations is not without its challenges. This is especially true for new users of the software. This article is intended to help new users understand the process and make suggestions for how to avoid common errors.
Data SGP is an important tool for educators to have at their disposal, but it is not a replacement for other forms of student assessments or data collection. In addition, educators should keep in mind that the SGP model is not a panacea and should be used as part of a broader approach to understanding student performance. Finally, the SGP model should be considered as a part of an ongoing effort to improve educational data and information systems.