How to Use OSPI Data SGPT
SGPT is a software application that uses longitudinal student assessment data to create statistical growth plots (SGPs). SGPs serve as an indicator of student progress when compared with academic peers and school districts/state education agencies use them as a way of monitoring whether students are making strides towards performance goals based on prior test scores/covariates.
With an SGP data set in hand, schools can generate growth reports to share with both students and parents. Unfortunately, implementing student growth percentiles requires significant time and resource investment – especially for districts already committed to other initiatives like data reporting. Furthermore, mathematical formulas used to calculate student growth percentiles can be complex, so finding appropriate training may prove challenging.
Data SGPs may also be more susceptible to spurious correlations than single point measurements due to difficulty in accurately predicting which students will remain in each group over an extended period. Furthermore, teachers aren’t guaranteed to assign themselves consistently each year resulting in inaccurate estimates of teacher effects; baseline-referenced SGPs can especially fall victim to this problem as cohorts of students often take five years before enrolling under one instructor.
Even with such reservations, implementation of SGPs remains a top priority for many school district leaders. Recognizing both their benefits and demands on staff and resources, school district leaders understand they must balance implementing new tools with existing demands on resources such as staffing. OSPI is committed to offering as much assistance as possible while districts implement SGPs; however given our limited funding and staffing we cannot always meet that need completely.
The SGPdata package includes both WIDE and LONG format data sets (sgpData and sgpData_LONG respectively), supported by lower level functions studentGrowthPercentiles and studentGrowthProjections; however, for most analyses you will likely benefit more from using higher level wrapper functions that leverage LONG data format as it offers several preparation and storage advantages.
To use the SGPdata package, you will require R – an open-source programming language available for free on Windows, Mac OSX and Linux computers. SGP analysis requires advanced R functions that may require familiarity with this software but there are numerous resources on CRAN which provide guidance for beginners or unfamiliar with it – in particular our Data Analysis Vignette provides this. Furthermore, should you have any inquiries about our package feel free to reach out by contacting us directly or posting questions directly into GitHub repository and we will do our best to respond as quickly as possible! If you would like updates when they become available please subscribe to our mailing list!
