Data SGP uses longitudinal student assessment data to generate statistical growth plots (SGPs) which illustrate relative student progress relative to academic peers. SGP estimates are computed using latent achievement trait models based on teacher evaluation criteria and student covariates; estimates can also be reduced through comparison of student SGPs to an identical cohort of similar-performing students at each year of measurement.
These SGP analyses often utilize multiple variables; however, the basic requirements include: an individual student identifier; date associated with each assessment event and associated assessment occurrence code as well as score values for all assessed items. SGP functions then convert these raw scores to scaled scores that can then be compared against an average scaled score across students in their grade and subject area in order to provide an indication of how close each student is meeting proficiency standards.
SGP uses historical growth trajectories of each student as well as projections for future years to estimate the likelihood that they will reach proficiency standards by the end of their educational career. These projections are calculated with latent achievement trait model estimates that take into account factors like prior test scores, academic covariates and teacher quality when making these projections.
This package not only offers standard SGP analyses, but it also includes high level functions called abcSGP and updateSGP that combine lower level functions into one function call to simplify operational analyses and source code related to them. Furthermore, an instructor number lookup table allows districts to assign instructors through test records of students.
A vignette provides instructions for using these functions to generate student growth and achievement plots. Furthermore, it details how to utilize sgptData_LONG: an anonymized panel dataset which contains 8 windows (3 windows annually) of assessment data in long format for three content areas.
Before beginning any analyses with SGP, it is crucial that you have an in-depth knowledge of its operation so you can trust in its results and resolve any potential issues efficiently. For first time users of this package, this step should especially important; documentation, vignettes and examples provide thorough explanations of calculations and processes involved with SGP analyses. If any questions about its operation arise that cannot be found here please reach out for help – the SGP team are eager to assist!