Data sgp leverages longitudinal student assessment data to produce statistical growth plots (SGP). These plots display the rate of progress for individual students and groups of students. They allow educators to identify areas of strength and weakness, guide instructional decisions, and track student achievement over time. SGPs also provide valuable information for students and families, enabling them to track student growth and success over the course of their education.
To create a SGP, we compare a student’s current performance to that of an identical cohort of students from an earlier year. The data we use to construct these comparisons come from up to five years of a student’s test score history, including the Badger Exam administered in 2014-15. The Badger Exam had a very different statewide performance profile than either the WKCE, which preceded it, or the Forward exam, which currently follows it.
We use a method called least squares regression to estimate the latent achievement trait model that generates each student’s current SGP. We then apply a series of statistical procedures to this model to determine how likely it is that the student’s current SGP will fall in a range defined by prior latent achievement and other covariates. SGP estimates can therefore be viewed as noisy measures of a student’s true growth.
These noisy measurements can be minimized by analyzing several years of compiled data from multiple cohorts in a single distribution. By doing so, any idiosyncratic patterns of one year’s scale score distribution can be smoothed out and the knots and boundaries that define SGP percentiles are more reliable over time.
The SGP percentiles displayed for each student are based on the distribution computed using these multiple years of accumulated data. The SGP percentiles are rounded to the nearest whole number and compared to the national average for the content area. The percentiles are also compared to the statewide median, providing an indication of whether a student is making adequate progress in relation to students from similar backgrounds and starting points.
The sgptData_LONG format contains 8 windows (3 windows each annually) of assessment data in LONG format for 3 content areas. This data set is used for all operational SGP analyses. It is recommended to use this data set for all your SGP analyses as it makes data management easier. Note that higher level SGP functions require the sgptData_LONG data set as well as a state specific meta-data file embedded in it. This is described in the SGP Functions Reference page.