Data SGP utilizes longitudinal student assessment data to create statistical growth plots (SGP), which measure students’ relative progress compared to their academic peers. SGPs serve as an accurate representation of students’ actual performance than do standardized test scores alone, yet creating them involves complex calculations with large margins for estimation errors.

As such, it is crucial that careful consideration be given when preparing student assessment data prior to conducting SGP analyses. A large portion of errors associated with using SGPs arise due to inaccurate or improper preparation. To assist users in this regard, the SGP package offers two exemplar WIDE and LONG formatted datasets (sgpData and sgpData_LONG respectively) along with wrapper functions such as studentGrowthPercentiles and studentGrowthProjections) which facilitate this task.

Step one of the Student Growth Projection process (SGP) involves compiling a longitudinal data set containing each students identifier, grade level/time association with assessment occurrences and numeric scores for each assessment occurrence. Step two entails calculating student growth percentiles and projections before merging scale scores back into master longitudinal record for merging. Finally graphical representations are produced.

SGP analyses require access to R software – available free on Windows, OSX and Linux computers – along with familiarity with its process. We suggest reading up on SGP package user guide or attending training session so as to become comfortable with this methodology before engaging any operational analyses supported by this package.

To utilize the SGP package, a user must have longitudinal student assessment data in either WIDE or LONG format. Lower level functions, studentGrowthPercentiles and studentGrowthProjections require WIDE formatted data while abcSGP and updateSGP offer increased preparation and storage benefits over WIDE data formats. If conducting any but simple analyses using SGP packages we advise formatting LONG data since most of their capabilities rely upon long data formats.

Once initial preparation steps have been completed, SGP analyses typically center around analyzing and interpreting results. This step should be straightforward and simple – many analyses we support employ a two step process: