Data SGP is a set of tools that leverage longitudinal student assessment data to produce statistical growth plots (SGPs). SGPs are an effective performance measure because they measure students’ relative progress against a growth standard established by their academic peers and the prior test scores that they have. However, creating SGPs from standardized test score data involves complex calculations and large estimation errors. This makes it difficult to accurately gauge whether a student is meeting an agreed upon growth target and, therefore, making sufficient progress towards their goals.
The goal of the data sgp package is to make the process of calculating and interpreting SGPs as simple as possible. We believe that most errors associated with SGP analyses stem back to issues related to the preparation of student assessment data and so we provide a range of functions for preparing data that can be used in SGP analysis, both at the lower level (studentGrowthPercentiles and studentGrowthProjections) and at the higher level (wrappers for the lower level functions).
SGP calculations are performed using two sets of data: a prior test score history and a current achievement profile. The SGPs that are estimated for each student are then compared to the prior SGPs of their academic peers in the same content area to determine how much progress they have made. The prior SGPs are also analyzed to identify the covariates that contribute to the differences in student achievement, and then the relationships between these covariates and the latent trait of student achievement are investigated.
The final step in estimating the SGP is to use these results to determine the growth standard that should be achieved by the student to reach their goal. The goal then becomes to identify how much the student must grow each year in order to meet their target by the end of their academic career, and to provide them with the necessary support to achieve this goal.
Data SGP helps teachers monitor the progress of their students and assess the effectiveness of their instructional practices. In addition, SGP allows administrators to develop and communicate clear performance goals for their students, which are based on the level of proficiency that is required for graduation and to identify those students who may require additional academic supports.
In addition to the basic SGP reports, the sgpData package includes a teacher lookup table (sgpData_INSTRUCTOR_NUMBER) that provides an anonymized list of instructors associated with each student’s test record for a given year. This can be useful if there is a need to identify the teacher who most directly contributed to a student’s SGP calculation. This information is particularly important if SGPs are being used to inform professional development. However, it is important to note that the relationships between latent traits and covariates that are identified in SGP models could be due to the fact that students from different backgrounds tend to be taught by teachers of varying quality. This is an issue that will be explored further in a later section.