The SGPdata package, installed when you install the SGP software, provides two common formats for longitudinal (time dependent) student assessment data: WIDE and LONG format. In WIDE format, each case/row represents a unique student and the columns represent variables associated with the student at different times. In LONG format, the time dependent information for each student is spread out across multiple rows in the data set. In general, you will be better off formatting your data in the LONG format if you plan on running SGP analyses operationally year after year, as there are numerous preparation and storage benefits to doing so.
The goal of SGP analysis is to identify latent achievement traits that are correlated with performance outcomes, such as grade or standardized test scores. To achieve this, SGP analyses utilize least squares regression modeling and Bayesian inference to estimate the underlying latent trait model, then compare the estimated results of each student against growth standards established via teacher evaluation criteria and student covariates. Errors in estimation can be minimized by comparing the estimates to baseline SGPs that are computed using data from an identical cohort of students at each point in time.
However, obtaining an adequate amount of baseline-referenced SGPs for every student in a school is often challenging and expensive. Additionally, correlations between baseline SGPs and prior year scale scores are unlikely to be exactly zero, introducing bias into interpretation of SGP results. For these reasons, many districts have switched to SGP analyses that compare current-year growth to scale scores computed from the same student population in the previous assessment cycle.
In addition to allowing districts to compute valid and meaningful student growth percentiles, SGP analysis allows districts to track students across multiple assessments. This functionality is especially useful for evaluating teacher quality and improving educator training programs. The sgpData_INSTRUCTOR_NUMBER field in the sgpData table is an invaluable lookup function that enables districts to link students with their instructors, as well as to make informed decisions about instructor compensation and professional development. Similarly, the SGP software allows districts to use sgpData to analyze the performance of groups of students who are grouped together by a particular factor, such as gender, race, or class. This approach can reveal a wide range of trends and patterns within the group that would not be possible with more conventional means, such as analyzing the results of each student’s individual tests.
