How to Analyze Data SGP
Data sgp is an important tool for teachers and administrators to use to measure student growth. SGP scores indicate a student’s relative growth compared to his or her academic peers in the same grade level for each content area assessed on a single test window. It is useful for assessing student progress over time and determining if students are growing more than, less than, or as much as expected.
SGP is also useful for comparing students to their academic peers across the state, district and country. SGP is calculated on a scale of 1-99, where higher numbers represent more relative student growth and lower numbers represent less relative student growth. The highest value in the SGP scale is 99, which represents a student who has grown more than 75 percent of his or her academic peers.
Using SGP requires several steps and can be conducted in two different ways. Typically, operational SGP analyses are conducted using the lower level functions studentGrowthPercentiles and studentGrowthProjections. These can be used with a wide or long format data set and can take up to five years of student assessment history to run.
In order to simplify the process of conducting SGP analyses, the SGP Package provides higher level functions prepareSGP and updateSGP that “wrap” the lower level functions and automate the necessary source code. These functions are often used by other R packages when calculating SGP, and they can also be called from within spreadsheet programs like Microsoft Excel.
We recommend the use of LONG formatted data sets for operational SGP analyses because the higher level SGP functions are designed to work with the long format. The use of long formatted data sets makes it much easier to manage the SGP analyses and to update them as new assessment data is available.
The sgpData_LONG formatted data set includes all of the SGP function variables required for running the student growth percentageiles and projections. It also includes an anonymized teacher-student lookup table, sgpData_INSTRUCTOR_NUMBER, that allows instructors to be associated with individual student test records. This information is useful when identifying one instructor responsible for most of the student’s learning, a critical step in measuring teacher or school effectiveness.
When analyzing SGP results, it is important to understand how the underlying calculations are done. For this reason, the sgpData_LONG formatted set of data includes detailed documentation on the SGP function variables and how to use them in conjunction with other R functions. It also contains example data to illustrate how the SGP functions work with the sgpData_LONG data set. In addition, the sgpData_LONG and sgpData_INSTRUCTOR_NUMBER data sets are available on our website for download in Microsoft Excel and open source format. The sgpData_LONG data file is a large file (500 MB) and should only be downloaded if you require the full dataset for your analysis. If you do not require the complete sgpData_LONG data, please contact the SGP support team for assistance. We may be able to provide an abbreviated version of the data for smaller downloads.
