Data SGP is a system that utilizes longitudinal student data to generate statistical growth plots (SGPs). An SGP displays the percentile rank of a current test score relative to prior achievement levels; higher percentile ranks indicate greater growth for that student. Teachers and administrators use SGPs to gauge whether students have grown more, less, or at an equal pace as their academic peers.
SGPs come in two formats: Window Specific SGPs compare and report student growth over specific time frames; Current SGPs provide the latest available SGP for each student as a check on progress over time. Both types are valuable tools for professional learning and reporting by teachers and administrators as well as for determining whether students have attained proficiency on subject matter tests.
An SGP measures student growth relative to others within his or her cohort on an overall scale from 1-99; higher numbers signify more relative progress while lower ones signal less. For instance, an SGP score of 75 indicates that a student has outscored approximately half of her or his classmates who share similar MCAS score histories on that test subject matter.
To compute Student Growth Profiles that compare students with their cohort, at least four years of test data is necessary. Experts who developed the SGP methodology suggest using at least this much data when computing baseline-referenced SGPs.
The sgptData_LONG dataset is an anonymized panel data set consisting of eight windows (3 windows annually) of assessment data in LONG format for three content areas – Early Literacy, Mathematics and Reading. Additionally, this dataset offers demographic/student categorization variables that can help create student aggregates through summarizeSGP function.
When performing SGP analyses with sgptData_LONG, seven variables are necessary: VALID_CASE, CONTENT_AREA, YEAR, ID, SCALE_SCORE, GRADE and ACHIEVEMENT_LEVEL (if running student growth projections). Note that two additional variables will only be required if specific output columns are selected for output analysis.
Based on your analysis needs, the sgptData_LONG data set may be formatted either WIDE or LONG formats. In general though, for anything but simple one-off analyses using higher level wrapper functions like studentGrowthPercentiles and studentGrowthProjections that use this format by default it may be better off formatting using LONG rather than WIDE formats; additionally LONG offers numerous preparation and storage advantages over WIDE formats.
Not only can users compute baseline-referenced SGPs, but a new feature now enables them to calculate cohort-referenced SGPs that condition on up to two prior year scale scores as well. This provides users with an invaluable capability: it enables them to assess student growth relative to other grade students who share an identical testing experience – an effective strategy for understanding student growth.