Student Growth Percentiles (SGP) are a measure of a student’s performance compared to their “academic peers.” Academic peers are students in the same grade who have similar MCAS scores from the previous year. This comparison is done using a statistical procedure called quantile regression that places students’ MCAS scores on a normative scale – the same scale that other students who have taken the assessment have performed on.
Student growth percentiles are reported in terms of how much a student has improved in their academic skills over time compared to their academic peers. SGPs are intended to be considered alongside scaled scores and achievement levels, in order to gain a complete picture of a student’s performance.
To find a student’s SGP, look for the number in the top right corner of their MCAS score report, next to a growth symbol. SGPs are reported for each individual MCAS test. Students can also be reported for a grouping of students, such as by gender, racial/ethnic identity, income status, or educational programs (e.g. sheltered English immersion, special education).
SGPs can be used to identify whether a student’s improvement over time is above or below the state average. They can also be used to identify the extent to which a student’s improvement over time is greater than or less than other students in their grade, school or district.
Because SGPs are calculated anew each year, differences in SGP rankings between years should be interpreted with caution. For example, a student who moves up from one decile to the next may have a very high SGP that does not mean they are performing well in the classroom.
While the lower level functions that do the calculations, sgptData_LONG and studentGrowthPercentiles require WIDE data, the higher level functions, such as createSGPs_LONG and createSGP_TABLE, work with LONG formatted data. In addition to SGPs, the higher level functions can generate Goodness of Fit figures for each grouping of students and an overall Goodness of Fit figure for each student aggregate.
While the data sgp package is designed to be easy to use, it does require familiarity with the programming language R. This is a free, open source software platform that can be run on Windows, OSX, or Linux and has a wide range of resources available for getting started. If you are not familiar with R, we recommend exploring the many available resources on CRAN, including those specifically for data analysis. The data sgp package is designed to make it easier for schools and districts to do more complex data analyses than those available in the R command line. However, even the most basic SGP analyses can be difficult to run without proper preparation of the data set. The most common errors encountered during SGP analyses revert back to these data preparation issues. Therefore, we strongly encourage users to follow the advice and guidance in the documentation for preparing your data. This will ensure that your analyses are as accurate as possible.