The data sgp package provides classes, functions, and data for analyzing growth percentiles and percentile growth projections/trajectories using large scale, longitudinal education assessment data. The package is written in the R software environment and can be run on any Windows, Linux, or OSX computer. Running any SGP analyses requires some familiarity with the software.
Data sgp uses the historical growth trajectories of Star examinees to predict what their potential future performance might look like, including whether they will reach/maintain proficiency and what they need to do in order to get there. The projections are updated regularly so that they are based on the most current data available.
Student growth percentiles are a measure of how much a student has improved in their academic skills over the course of an assessment year as compared to their “academic peers.” A student’s academic peer group is comprised of students who have similar score histories in their same grade and subject area.
The SGP calculation compares a student’s score to those of the average (median) of their academic peers. Students with higher than average scores in their grade and subject area can experience varied growth from year to year. This is why it is important to consider a student’s overall improvement in their academic skills and not just focus on their individual test scores.
A student’s SGP is calculated from the results of their most recent state assessments and at least one prior assessment from an earlier testing window. In the Star Growth Report, the SGP for a student is reported based on either the most recent testing window or the selected prior school year during the report customization process (in the Timeframe drop-down list).
For teachers, a median student growth percentile (mSGP) is derived from the average of the teacher’s mSGPs in their grade and subject area over the past two years. This is a more accurate measure of the teacher’s actual effectiveness than is a mean mSGP because it accounts for the variability in student growth exhibited by even very high performers.
Teachers are rewarded for their teaching abilities by earning an mSGP in their grade and subject area in Fall of each year, assuming they have at least two years of mSGP data. In 2023-2024, teachers who have two years of mSGP data will earn their scores based on the most recent or the median of their previous mSGPs.
sgpData is an anonymized, panel data set comprising five years of annual, vertically scaled, assessment data in WIDE format. This exemplar data set models the format of the data used with the lower level SGP functions, studentGrowthPercentiles and studentGrowthProjections. The first column in the sgpData table, ID, provides each student with a unique identifier. The next 5 columns provide the assessment occurrences for each student, including the student’s assessment score in each of the content areas. The last column, sgpData_INSTRUCTOR_NUMBER, provides the insturctor associated with each student’s assessment record. The sgpData_INSTRUCTOR_NUMBER field is not mandatory, but it can be useful for tracking teacher-student relationships.