Data sgp leverages longitudinal student assessment data to produce statistical growth plots (SGP) that measure students’ relative progress compared with academic peers. These are a critical tool in teacher evaluation systems because they provide insight into whether students’ achievement levels meet or exceed established growth standards based upon student covariates. However, constructing SGPs from students’ standardized test score histories involves complex estimation algorithms that can introduce substantial error. The purpose of data sgp is to provide district educators with SGP estimates that are as accurate as possible given the limitations of these methodologies.
Using state-specific growth percentiles, we compare the performance of current year test takers with those in previous years to calculate SGPs. The more the student scores above the mean, the higher the SGP. Teachers and districts can see their individual students’ SGPs by selecting a student in the Star Growth Report and choosing the “SGP Data” tab. The SGP data spreadsheet displays the SGP of the selected student, including a five year history of SGPs for each subject.
We also use SGPs to help teachers and administrators understand the range of student growth trajectories possible from one window of assessment data to the next. This information is used to support teacher planning, inform classroom practices, and evaluate schools/districts.
Unlike medians, SGPs are more sensitive to classroom composition than VAMs because they do not mask differences in teacher effectiveness. Moreover, because SGPs are based on students’ trajectories of learning, they are more useful than absolute student growth measures such as a student’s percentile rank.
This vignette describes the process of creating SGPs from a student’s standardized test scores over time, using SGP analyses implemented in the rOpenSci package. While the focus of this vignette is on SGPs, similar procedures can be applied to any other type of student outcome measure, such as graduation rates or test score gains over time.
The SGP vignette includes a step by step guide to analyzing data and interpreting the results using both graphical and textual methods. It is a great starting point for any district interested in conducting SGPs.
SGP analyses are most efficiently run on LONG formatted data sets. The sgptData_LONG dataset provides an anonymized panel of LONG formatted assessments in 8 windows (3 windows annually) for three content areas (Early Literacy, Mathematics and Reading). It contains all necessary data for SGP analyses and also includes the sgpData_INSTRUCTOR_NUMBER lookup table, which allows users to associate instructors with student assessment records via unique identifiers.