What Is Data SGP?
Data sgp are statistical values that measure students’ progress toward meeting grade-level academic content standards on state assessments. They are calculated by comparing each student’s current assessment score to his or her performance relative to the performance of academic peers with similar assessment score histories. This process allows for a more accurate comparison of growth than simply evaluating changes in scale scores between students, because it takes into account the degree of difficulty of each assessment and the differing grade-levels of the tests used to calculate a student’s scale score.
Student Growth Percentiles (SGP) are based on a comparison of students’ MCAS test score histories. A student’s SGP indicates that he or she scored the same as or higher than students with similar academic peer groups from previous MCAS administrations for a given subject area. Academic peer groups can include students in the same grade with similar MCAS scores, but they can also be students from different grades or academic peer groups defined by such characteristics as race/ethnicity, special education status or multilingual learning.
A student’s SGP is a percentile value, so it can range from 1 to 99. A student with an 85 SGP means that he or she showed more growth than 85 percent of his or her academic peers in that subject area during the course of one year. For this reason, SGPs are useful for students with a wide range of achievement levels, from low to high performers.
SGPs can be reported at the student level, by classroom, school and district, but they can also be analyzed at the system level. In addition to using SGPs to identify and celebrate student success, they can be used to identify areas of opportunity for school improvement and support efforts to reach proficiency for all students.
As a custom, relational database, the SGP package is designed to be as user friendly as possible. However, since it is a tool for statistical analyses of longitudinal data, the SGP package requires access to computers running the R software. R is freely available for Windows, OSX and Linux and there are many resources on CRAN for users new to the program.
The SGP package provides a sample WIDE format data set, sgpData, that can be used to run SGP analyses. This data set includes a student-level table that provides information for each student, including his or her ID and the teachers assigned to the student for each of the five years of statewide testing in which they were assessed. A teacher-level table is also included in the sgpData dataset, providing information about each of the teachers who assessed the student for each year in which they were tested. For this reason, the SGP analysis vignette assumes that you will use the sgpData exemplar for your SGP analyses. However, you can use any WIDE format data set you have available for your analyses as long as it contains the required columns. The sgpData vignette also discusses the use of LONG format data sets and provides a link to a more comprehensive documentation for working with LONG format data.