Data sgp is an analysis tool for educational assessment data designed to run on the open source statistical software environment, R. Like any data analysis the bulk of the time required to conduct SGP analyses is spent in data preparation; errors in SGP calculations are often due to incorrect or missing data elements. Once prepared correctly SGP analyses are relatively simple and straightforward.
Unlike many assessment metrics SGPs are not based on absolute values but rather on the relative performance of academic peers; a student’s growth percentile indicates how much their raw score on a test section has grown in comparison to that of their academic peers and can be used by educators/administrators to help determine how much progress a student needs to make to reach proficiency. Growth percentiles can be used in a number of ways including identifying low achieving students, informing instructional practice, supporting classroom research initiatives as well as evaluating schools/districts.
In addition to providing educators with a clear picture of student progress SGP also helps in communicating to parents/communities that progress toward achievement targets is necessary despite competing priorities. Using SGP allows districts to establish multi-year growth standards based on official state achievement goals/targets rather than having to rely solely on a single grade level test score.
At the state level median SGPs are nearly always 50 (assuming norms were established using only current year test scores) but this is due to both the fact that half of the states have growth below the national median and the use of a single test in establishing growth percentages. In order to more accurately represent the true range of potential SGPs for a given student it is necessary to include two tests from different testing windows.
To do this the SGP package offers a lower level data set called sgptData_LONG and higher level wrapper functions for creating student growth percentages and projections called studentGrowthPercentiles and studentGrowthProjections; sgptData_LONG includes the 7 required variables (VALID_CASE, CONTENT_AREA, YEAR, ID, SCALE_SCORE, GRADE, ACHIEVEMENT_LEVEL) as well as the data element sgpSummary. For those looking to run SGP analyses operationally on an ongoing basis it is recommended to utilize LONG format student assessment data as this has numerous preparation and storage benefits compared to WIDE data sets.
sgpData_LONG is a sample data set of 8 windows (3 windows annually) of assessment data in LONG format for 3 content areas that can be used to demonstrate how the SGP package works and the types of analyses that are possible. For those interested in experimenting with this software and developing their own SGP analyses there are numerous online resources available to help get started.