Data sgp is a software package used to calculate student growth percentiles and projections/trajectories using large scale, longitudinal education assessment data such as from standardized tests, portfolios or other sources. The package allows for numerous analyses including comparisons across time; effectiveness evaluation for interventions or policies as well as educational policy impact analyses.

SGPs provide a more complete picture of student achievement than simple average test scores do, as they reflect both the extent to which a program has worked as well as any improvements that students have experienced over time. It should be noted, however, that SGPs do not fully represent academic performance – they’re only one factor among many and do not account for factors like family income or afterschool programs that may affect a child’s success.

These analyses can provide educators and other stakeholders with important data for making decisions to enhance quality instruction, curriculum development, school facilities and other aspects of educational systems. Furthermore, their results can help evaluate program effectiveness by highlighting any that do not fulfill their intended goals.

However, unlike most analyses of student assessment data which employ similar statistical techniques to both produce SGPs and their estimates of covariates simultaneously, our approach employs an additional method which isolates the true covariate effects independent of how they are measured – this allows us to interpret and understand each student covariate’s impact despite any inaccuracies in its estimate.

We conducted extensive performance comparisons of various approaches and discovered that our model is the most accurate in terms of both its ability to produce realistically estimated expected true SGPs, and its sensitivity to covariate effects underlying actual analyses and methods. This finding is highly significant given that reliability of SGP estimates can differ between analyses or methods and can have a dramatic impact on conclusions drawn from them.

Installing data sgp includes the SGPdata package, which provides sample WIDE and LONG formatted datasets to illustrate its functions. LONG format is generally preferred as it offers numerous preparation and storage benefits over WIDE formats of data.

SGPdata package can be downloaded at https://github.com/sgp/data/releases and for Windows is available from Microsoft repository. Both software packages are free and open source with support from an international community of users; when applying it to their own data users should be aware of its limitations and seek expert advice before doing so. A minimum memory requirement of 4 GB for long format analyses and 2GB is needed for wide format analyses respectively – please see SGP package documentation for details.