The data sgp package is designed to run student growth percentile (SGP) analyses using the R programming environment. R is free and open source software that runs on Windows, OSX and Linux. Running SGP analyses requires familiarity with the R programing language. There are many resources available on the internet to help get you started with R.
SGPs are based on comparing students with similar score histories and determining relative performance. They are calculated using up to two years of historical MCAS test data. This data is used to identify academic peers for each student and then use a statistical procedure called quantile regression to place their current scores on a normative scale, making it possible to identify the percentile rank of their performance. Academic peers are derived from all students statewide in the same grade and include all demographic groups, including students participating in educational programs (e.g. sheltered English immersion, special education).
Students who have been in the same classroom for at least 70% of the school year before taking the state assessment can be included in the SGP calculation. The teacher of record must be in the class for at least 20% of the school year before the state assessment in order to have a valid SGP estimate. The teacher of record must also have a sufficient number of SGP estimates in the same class to create a reliable average. The data sgp package provides an easy way to calculate and display these estimates for teachers with valid students.
In addition to SGP calculations and reports, the data sgp package also provides teacher level aggregates. These aggregates can be sorted by a variety of factors and used to make decisions about teaching practice and professional development. The aggregates can also be used to monitor trends over time.
The data sgp package contains 4 examplar data sets for use with SGP analyses. The first, sgpData, specifies the data in the WIDE format that’s used with the lower level SGP functions studentGrowthPercentiles and studentGrowthProjections. The next two, sgptData_LONG and sgpData_INSTRUCTOR_NUMBER, specify data in the LONG format used by higher level SGP functions like abcSGP, prepareSGP and analyzeSGP. These data sets contain 7 required variables: VALID_CASE, CONTENT_AREA, YEAR, ID, SCALE_SCORE, GRADE and ACHIEVEMENT_LEVEL. sgptData_LONG also includes an anonymized instructor-student lookup table that is utilized to produce teacher level aggregates.