The data sgp dataset contains assessment scores in long format for five years for each student. The first column, ID, provides the student unique identifier and subsequent columns provide the score for each assessment in the year indicated. These scores are then standardized and aggregated to generate student growth reports. This data is available to all districts, schools and individuals in the state of Massachusetts for the purposes of evaluating student performance and supporting instructional practices.
While the calculations involved in SGP are complex, the information produced can be easily shared in terms that are familiar to teachers and parents. SGPs describe relative growth on a scale of 1-99, with lower numbers indicating less relative growth and higher numbers indicating greater growth. For example, a student who receives an SGP of 50 has demonstrated growth that is equal to or greater than half of students with similar prior test scores (their academic peers).
SGP analyses can be conducted for both windows and current data. For windows, SGPs are calculated based on the most recent assessment in the window and up to two previous assessments in the same testing window. The default choice in the sgp package for SGP analysis is to use the last assessment in the window and one prior assessment. Current SGPs are calculated for students that have taken a single assessment in the past 18 months, and if possible, two assessments from different testing windows.
Both of these analyses are used in the SGP dashboard to help educators track students’ learning over time. The dashboard also provides a visual depiction of student-level, grade-level, and group performance, including comparisons across groups of interest. In general, SGP analyses are designed to be as simple as possible and any errors that may arise usually revert back to problems with data preparation.
When a program accelerates too quickly, the SGP for the middle student can drop significantly, leaving only the slowest and fastest students making progress in class. This situation is most often encountered in math programs, where students are placed into different groups based on their previous grade-level performance. SGPs allow teachers to identify which students are not keeping pace with their classmates and are falling behind. This can be an indication that the accelerated instruction is not appropriate for those students. In such cases, the program should be revised so that a more appropriate level of acceleration is applied to those students. This should also be communicated to the entire school community. Read more in the SGP data analysis vignette for detailed guidance on using the SGP data analysis tools.