Data SGP is a system that utilizes longitudinal student assessment data to produce statistical growth plots (SGPs) that allow teachers to see a student’s current test score in the context of other students with similar prior achievement levels. SGPs are more accurate than unadjusted scores and provide a more holistic picture of a student’s performance by considering all of the stages in his or her academic career.
The sgpData package contains several percentile rank tables that outline a student’s performance relative to his or her academic peers. These percentile rank tables are helpful in identifying areas where students may be struggling, and in starting discussions between curriculum coordinators or principals about how to improve instruction.
Percentile rank tables are available for students in grades 1-8 in all content areas, except science (which is only administered in grade 9). These tables show the average test score for each student across five years of assessment and the percentage of the total test score that was scored at or above the respective percentile.
The data sgp tabel also provides a number of other statistical indicators that are useful for understanding student performance and progress, such as mean and median standard deviation scores. These statistics, when combined with other data sgp, can help identify patterns in test results that may point to a lack of knowledge or ineffective instructional practices.
To view these data sgp tables, users can log in to their Star account and click on the “Datasets” tab. They can then select the district and school level for which they would like to see the data. Data sets are available in both WIDE and LONG format. The SGP data set sgpData, installed as part of the data sgp package, includes exemplar WIDE and LONG format datasets to help get started with data sgp analyses.
In addition to the standard percentile rank tables, data sgp includes an instructor number lookup table called sgpData_INSTRUCTOR_NUMBER that displays the instructor numbers associated with each students’ test records. This information is helpful in determining how many teachers are responsible for a particular student’s academic development over time, and can help determine whether certain instructors are more effective at guiding their students to success than others.
The sgpData package also includes lower level functions, studentGrowthPercentiles and studentGrowthProjections, that perform the same analyses as these higher level functions but with much less code. In most operational SGP analyses, the higher level functions are used together to minimize source code. Additionally, the SGP package provides wrapper functions abcSGP and updateSGP that simplify the code required to conduct a SGP analysis. The following example shows how to use these wrapper functions to create a student growth plot using the sgpData data set.