Data Sgp is a package consisting of classes, functions and data sets designed to analyze student growth percentiles using longitudinal educational assessment data. Using quantitativeile regression techniques it estimates conditional density estimates on achievement history data while also producing projections (known as percentile growth trajectories) showing what level of growth will enable a student to reach future achievement goals.
Employing longitudinal data for SGP analyses requires using a WIDE format. WIDE formats organize cases/rows as unique students while columns represent time dependent variables associated with them. SGPdata comes preloaded with SGP package to assist users when working with WIDE data sets such as sgpData and sgpData_LONG for ease of use with this application.
SGP utilizes historical growth trajectories of Star examinees to project what level of growth is necessary for an individual student to reach proficiency on an assessment within a reasonable timeline. The SGP report includes potential trajectories available to each student given current performance levels and district screening windows specified during customization of SGP reports (via Window Specific SGPs section).
Below is an illustration of a student growth trajectory. The vertical axis represents their current score while the horizontal axis indicates their required amount of growth to reach proficiency. Color indicates an estimated probability that students will reach the right side of the graph; that is, percentages who grow past this point are indicated on this plot.
In this example, a student is projected to achieve a score of 70 on their next assessment test, meaning that in order to reach this goal they need to outperform 85 percent of their peers academically. Unfortunately, several factors impede a student’s growth and ability to reach their proficiency target, including available resources and motivation levels. As such, it is crucial that educators thoroughly analyze student profiles and consider all factors when creating growth plans for them. This article serves as an excellent source for more information regarding growth percentiles and setting individual student growth targets. Adam Van Iwaarden wrote this article which covers various student growth plans and what factors must be taken into consideration when creating them. You can access his article at this link. An accomplished developer of statistical software, currently employed at the University of Colorado. For over ten years he has worked extensively in statistics, machine learning and education technology fields. He has contributed to a variety of open source projects, including the statistical R package SGP. He has written for Journal of Statistical Software, PLoS One and RStudio publications as well as serving on the board for EdCamp Denver non-profit educational organization. He is an active contributor of both projects on GitHub while serving on board of directors of EdCamp Denver non-profit organization.