Student growth percentiles provide an indicator of how much a student has improved relative to his or her academic peers. Calculations involve using statistical methods which compare assessment scores against peers’ assessments scores; percentages may also reflect normative data that suggests those who typically perform better are expected to increase more rapidly in performance.
Students are evaluated against their academic peers, which may include those in their same grade and demographic groupings such as race/ethnicity, special education needs or multilingual learning. Students can also be organized into academic cohorts who share similar score histories on state assessments – these cohorts could be defined by school district or entire state.
Student growth percentiles are determined based on students’ performance on state assessments. Analysis results are used to create a predicted percentile for that student that represents how many points they should expect to improve over the coming year in order to meet academic goals for next year – this predicted percentile can then be compared with their actual percentile achieved on state assessments to generate their growth percentile.
Student growth percentiles provide a more accurate representation of a student’s current performance than traditional assessment reports do. Student growth percentiles measure how much an individual student has improved on assessments relative to his or her peers and can identify both areas of strength and weakness within an academic subject matter area.
As a rule, we recommend the LONG format when managing data for SGP analyses. It is more straightforward than managing wide formats, and most higher level functions available within SGP were specifically developed to run on LONG data sets. Furthermore, many of these higher level functions assume that both embedded sgptData_LONG and INSTRUCTOR-STUDENT lookup files already exist in an SGP data set, streamlining operational analyses.
Note that SGPs are calculated anew each year, so any differences in student percentile rankings from one year to the next should be treated with extreme caution. As a general guideline, any variations of 10 points or less between years should be viewed as significant. When comparing SGPs between schools it should also be done cautiously as differences could reflect influences that are unique to each institution and do not correlate directly to student achievement.