The data sgp package contains classes, functions and data for analyzing large scale, longitudinal education assessment data. Its lower level functions studentGrowthPercentiles and studentGrowthProjections perform calculations such as conditional density estimation and derived coefficient matrix calculation that estimate percentile growth projections/trajectories for individual students based on their current achievement history and the estimated needed percentile growth to reach desired future performance targets.
SGP measures relative student growth by comparing the assessment scores of an individual student to the assessments scores of similarly academically-similar students. By understanding an individual student’s relative performance compared to peers, educators can make informed decisions regarding a student’s educational needs and provide appropriate support for the student.
To run SGP analyses, a user will need the free and open-source programming language R installed on their computer. The SGP package also requires a set of additional libraries to be loaded. Several online resources exist to assist with the installation and setup of the R software for those who are unfamiliar. SGP analysis also makes use of advanced R functions that require familiarity with their use – see the SGP Data Analysis Vignette for more details.
In order to utilize the SGP functionality, a data set in WIDE format is required. The sgpData data set contains an anonymized panel data set of 5 years of vertically scaled assessment results in the WIDE format. The first column provides the unique student identifier and the next five columns (SS_2013, SS_2014, SS_2015, SS_2016 and SS_2017) contain the students assessment scores at each of these grade levels over time.
To use the SGP functions, a user will need to load the sgpData data set and sgpData_INSTRUCTOR_NUMBER, an anonymized lookup table that provides the instuctor information associated with each students test record. The sgpData_INSTRUCTOR_NUMBER is used to identify which teachers an individual student was assessed by in each of the content areas for a given year. The sgpData_INSTRUCTOR_NUMBER can be loaded multiple times for each student if the educator taught the student in more than one content area for a given year. The SGP data is sorted by relative student growth percentile so that the most recent percentile is listed at the top of the list and the lowest percentage of relative growth is at the bottom of the list. The SGP percentile range is 1-99, with higher numbers indicating more relative student growth. A SGP of 50 indicates that a student demonstrated growth in the subject-matter content area equal to or greater than half of the students with comparable score histories on that particular subject-matter test. A SGP of 100 would indicate that the student is growing at or above the state average. If the SGP percentile is below the state average, then it indicates that the student is not demonstrating enough growth to be on track for success in that subject area. This information is helpful for guiding instructional strategies to improve student outcomes. The resulting SGP report is displayed in the student profile page.