Data sgp is a set of classes, functions and datasets that use large scale, longitudinal education assessment data to estimate student growth percentiles and project/project student achievement trajectories. These calculations are based on a process called quantile regression, where conditional density is estimated for each student’s prior and current achievement scores and then used to “encode” the relationship between these two sets of values, allowing us to qualify a student’s level of growth and predict what different rates of growth will yield for students statewide in terms of their likelihood of meeting proficiency standards in any given future year.
The sgpData dataset is an anonymized, panel data set that provides exemplar longitudinal student assessment data in Wide format. It includes 5 years of annual, vertically scaled, test data, including student and teacher information associated with each assessment occurrence for a single content area. It serves as a model for the sgpData_LONG and sgpData_INSTRUCTOR lookup files, used in the SGP functions to prepare SGP data and generate SGP results.
Each year, DESE calculates SGPs for all students in grades 4 through 8 and grade 10. SGPs in ELA and mathematics compare a student’s previous test score with the test scores of other students who have taken the same grade-level MCAS tests. Growth trajectories are not available for grade 3 and science (because of the lack of a comparable prior test).
SGP analyses convert raw student scores into scaled form, which can be compared to an average of the scaled scores for all students in the same grade and subject area. By doing so, SGP can identify whether a student’s scaled score falls above, below or at the same level as this average. It can also indicate how much more or less than other students in the same grade and subject area need to grow to reach proficiency in that subject area.
By using SGPs, educators can get a better picture of their student’s progress and understand how they may need to focus instruction to help all students make meaningful gains in academic achievement. Individual student SGPs can be shared with parents, providing them with a richer understanding of their child’s progress. At the classroom level, educators can incorporate student SGPs into their own set of Student Learning Outcomes (SLOs), and administrators can use SGPs to guide district-wide improvement efforts.
The SGP functions in the data sgp package are designed to perform the following 6 steps: