A key goal of data sgp is to allow researchers to access, analyze and compare the data sets that are available for specific research questions. This is accomplished by using common terms and formats for the data sets that are collected and stored. This ensures that the results from one analysis can be compared to the results of another without having to worry about differences in terms, definitions or formatting that may introduce bias.

In addition to allowing direct comparisons of the data sets, data sgp also allows researchers to use different methods for analyzing the same data set. For example, data sgp can be used to calculate student growth percentiles as well as standard normal distributions of student assessment scores. This flexibility provides an additional layer of transparency that is useful for researchers and decision makers alike.

Data SGP is an open source software package that runs on a computer with the R statistical programming environment. R is a free, cross-platform application that can be run on Windows, Mac OSX and Linux. In order to run SGP analyses, the user must have at least a quad core machine that is capable of running R with 4GB of RAM allocated to the application for each analysis run.

The SGP package installs exemplar WIDE and LONG data sets that are used for demonstrating various SGP functions. A vignette on SGPdata provides further detail on the format of these data sets. In short, a case/row in a data set represents a single student and the variables associated with that student at different points in time. The sgpData_INSTRUCTOR_NUMBER data set is an anonymized student-instructor lookup table that associates teacher information with each students test record. A student can have multiple teachers associated with their test records for a single content area in a given year.

In order to run student growth percentiles, the user must have the test-specific Conditional Standard Errors of Measurement (CSEM) for each state test available for simulation. These CSEMs are typically obtained from the NSPS website in a spreadsheet file. The sgpSGP_COUPLED argument passed to the studentGrowthPercentiles function indicates whether or not the CSEMs are to be combined for the calculation of confidence intervals. This is set to TRUE by default.

The ability to run SGP analyses using real-time state data is a major factor that motivates us to continue to develop and maintain the SGP package. We hope that others will find it to be a valuable resource as they continue their research and make decisions about educational policy. We would like to encourage researchers and decision makers to contact us with ideas about additional features that they would like to see in the future. In addition, if you have a problem with the SGP package, please do not hesitate to contact us by writing a message or opening an issue on GitHub. We will do our best to assist you. Thank you for your support of the SGP project!