May 16, 2024

Data SGP – Getting Started With Data SGP

Data SGP is a statistical software package for the R program environment. To use it one needs a computer running Windows, OSX, or Linux that has R installed (or can access the free and open source version of it). Running SGP analyses is intended to be a two step process – data preparation and analysis. We provide detailed instructions for both steps of the process on our help pages and on the GitHub page where the code is located.

We also have a number of additional online resources for getting started with the SGP program including an online video tutorial, detailed documentation of all functions in the program, and a comprehensive FAQ. We encourage anyone with questions to get in touch and to explore the various online resources.

A key feature of our SGP analysis tools is the ability to compare and report student growth percentiles over time. This provides teachers with a powerful tool for understanding and communicating student growth results. Students are ranked relative to their peers and higher numbers indicate more growth than students with lower scores.

This information can be displayed using the standard graphing features of the SGP program. Alternatively, it can be summarized with an SGP table that combines and reports the most relevant information for each student. This is especially useful for comparing individual students.

SGP tables can be created and manipulated using the function prepareSGP(). These tables are then used by the lower level SGP functions studentGrowthPercentiles and studentGrowthProjections to produce teacher and student aggregates.

The sgpdata package contains 4 examplar data sets to assist with SGP analyses. The first, sgpData, specifies data in the WIDE format that is used by the lower level SGP functions. The second, sgptData_LONG, specifies data in the LONG format used by the higher level SGP functions summarizeSGP and analyzeSGP. The final sgptData_INSTRUCTOR_NUMBER data set is an instructor-student lookup table that is utilized to produce teacher level aggregates.

In addition to the basic information contained in these tables, there are some specific variables that are required if you want to run an SGP analysis. These are the VALID_CASE, CONTENT_AREA, YEAR, ID, SCALE_SCORE, GRADE and ACHIEVEMENT_LEVEL variables. You may also wish to include a column for student first and last name if you plan on creating individual student growth or achievement plots.

The most common uses of SGP are for student-level comparisons and for assessing individual students. Compared to standard graphs, SGP plots have a more graphical and intuitive way of presenting comparative information. They also provide a more meaningful and accurate assessment of student growth by eliminating the need to calculate an average or summing scores for each student.

In order to utilize this type of information, teachers need access to a high quality data set. Baseline-referenced SGPs require at least three years of consistent and stable assessment data to produce models that can be compared with scale scores from a prior year. These requirements limit the number of instructors in a given cohort that can be evaluated based on their SGPs and make them more vulnerable to spurious correlations.