Student Growth Percentiles (SGP) are a statistical measure of how well students have learned over time and are progressing toward proficiency. They are calculated by comparing the performance of a student to that of academic peers across the state who have similar score histories.

While a student’s SGP may vary from year to year, the overall trend for any given grade is indicative of whether the student is moving closer to or further away from meeting the state learning standards in their particular content area. This information should be shared with teachers, parents and community members, as it can provide additional insight into a student’s progress and achievement.

SGPs are calculated by analyzing up to two years of historical MCAS data for a single student. These data are normalized to a common scale – 0-100, which can be used to compare performance between different years and assessments.

Typically, a student’s SGP is a number between 1 and 99 indicating how much they have grown compared to their academic peers. Teachers can identify student SGPs in their classroom, which are displayed in the teacher growth graph, and use them to guide planning and instruction. Districts can review SGPs at the school and district levels, and incorporate them into their educators’ and administrators’ standards-based system of evaluation.

To calculate SGPs, historical MCAS data is analyzed using a statistical technique called quantile regression. The data are normalized to a common scale and then grouped by academic peer groups based on similar test history and demographic characteristics, such as gender, race/ethnicity, educational program (e.g., sheltered English immersion), and school type. Academic peers are grouped so that students who scored the same on the same test receive the same SGP.

When SGPs are compared to each other, the higher number indicates more relative growth. For example, a student with an SGP of 75 has demonstrated growth that is greater than half of their academic peers.

As more data are collected and analyzed, average SGPs can be produced for schools, districts, and subgroups of students within schools or districts. However, the average SGP for a specific school, district, or group will fluctuate because it only includes a small sample of the available data.

There are several variables that can be adjusted for SGP analyses – the number of panel years to use for projections and lagged projections, the state associated with the data for access to embedded knots and boundaries, cutscores, and CSEMs, and the baseline coefficient matrices used for projections (when available in SGPstateData – this is very computationally intensive). The best option is to run the lower level functions, studentGrowthPercentiles and studentGrowthProjections, with as much data prepared as possible. This will maximize the ability to identify patterns in student and teacher performance and to predict future test results.