A data sgp is a set of aggregated student performance measures collected over time that teachers and administrators use to make decisions about instruction and assessment. These measures can include individual-level student performance metrics like test scores and growth percentiles, as well as school and district-level measures like class size and attendance rates. Data sgp can be used to identify areas for improvement, inform classroom practices, evaluate schools and districts, and support broader research initiatives.
SGP measures student test score progression over time by comparing a student’s raw score on one assessment to the mean (or average) score of students with similar prior test scores (their academic peers). SGP is reported as a percentage and can be easily compared to state benchmarks to provide educators with a clear picture of their students’ growth. This information can be valuable to teachers and parents as they assess their own students’ progress, identify high performing and underperforming students, and track the impact of instructional interventions.
In addition to the SGP summary report, educators can access a more detailed SGP data spreadsheet by selecting a student in the report and choosing the “SGP Data” tab. This spreadsheet displays a student’s SGP results for each of five testing windows. The first column, ID, provides a unique identifier for each student; the remaining five columns, SS_2013, SS_2014, SS_2015, and SS_2016, provide the student’s assessment results from each of these years.
The sgpData_INSTRUCTOR_NUMBER table is an anonymized, teacher-instructor lookup table that provides the instructor number associated with each student’s test record. This is a key piece of the data set because it allows for comparisons between students with different teachers, and helps to ensure that SGP estimates are valid and reliable.
Michigan uses student test score progression data in educator evaluation systems; therefore, being able to accurately and efficiently compare students and teachers across instructors is critical. SGP provides this functionality in a way that is not possible using standard growth models or other methods.
To be effective, data sgp must have many things in common with other datasets such as demographic and student background characteristics. Data sgp must be collected, cleaned, and processed before it can be used to make comparisons between groups of students, or between teachers or schools. This process is called preprocessing and is a critical step in making sure that SGP estimates are valid and reliable. Preprocessing removes covariances between the test score predictors and student-level variables that affect growth model estimations, helping to reduce estimation error and improve consistency between different SGP estimates. Data sgp is still a work in progress, but we are working hard to collect and create this data set, along with analytical tools for interpreting it. Despite the hype surrounding the term ‘big data’, the amount of data sgp we are collecting is relatively modest in comparison to other large datasets such as global satellite imagery and global Facebook interactions. Nonetheless, the effort to collect, process and analyze this data set will be significant.
