Data Sgp is a collection of tools and features that allows dynamically uploading of data into knowledge bases. While supporting standard formats for uploading data, Data Sgp allows for the use of custom schema for managing personal datasets. Furthermore, it offers a view of all training datasets which can be used for fine-tuning models; you may manually upload new training datasets as needed or there may already be available training datasets provided by Scale AI depending on your installation.
Student growth percentiles (SGPs) show you the extent to which your students have improved on state assessments relative to their academic peers, regardless of subject matter or administration in spring or fall. SGPs are calculated using quantile regression – an effective statistical technique which accounts for differences in scale scores across years as they change. These methods ensure that a student’s SGP remains constant during assessment transitions so you can easily compare performance over time.
An SGP indicates how much a student’s performance improved compared to that of his or her academic peers on any MCAS test. Academic peers are identified based on MCAS score histories of students from across Massachusetts who share similar demographic characteristics (gender, race/ethnic background and educational program participation e.g. sheltered English immersion or special education).
As SGPs are determined by comparing students against their academic peers, they will always have some correlation with proficiency rates; however, any such correlation should be seen merely as a statistical artifact and not evidence of direct causality between growth and proficiency rates.
The SGP model is regularly updated based on historical growth trajectories of past Star examinees. Projections generated by this model are then applied to current grade levels in order to estimate how many points additional points a student will need in order to attain proficiency on future Star administrations; these estimates can then be displayed in the Star Growth Report, and updated as more data become available.
SGPs are reported directly to students and parents via Individual Student Reports (ISR) for both Math and ELA. Schools and districts also receive SGP reports via NJ SMART’s mSGP Reports which display similar information but also provide additional analysis regarding progress toward proficiency.
In addition to showing a student’s SGP, the mSGP Reports display their growth by subject and grade level. Educators can use these reports as an insight tool in order to identify academic areas where additional support may be necessary; then use that insight to guide instructional decisions and drive student achievement.