Data Sgp – Using Data Sgp to Identify Areas for Improvement

data sgp

Data Sgp is an invaluable resource that enables teachers to identify areas for improvement in their instruction, narrow the gap between high and low performers in school settings and ensure resources reach those most in need more quickly and efficiently. Additionally, Data Sgp ensures students receive top quality education taught by qualified instructors.

Executing SGP analyses requires a computer capable of running the open source statistical software R, available free from its official R Project website and easily downloaded and installed. Once properly set up, SGP analyses become straightforward – with most errors encountered due to data preparation issues being quickly addressed, so most time spent performing SGP analyses being used up by this task alone.

The sgpData_WIDE file is an anonymized panel data set composed of five years’ annual, vertically scaled assessments for every student over five years. Rows provide each unique identifier while columns reflect time dependent variables associated with assessment occurrences for that student. This exemplary data set serves as a template for use with studentGrowthPercentiles and studentGrowthProjections lower level functions.

SGP analyses are intended to be operationally efficient; most calculations are completed by lower level functions while higher level wrapper functions, like summarizeSGP and studentGrowthProjections, only require raw data in order to aggregate or plot. Therefore, all higher level functions are compatible with both wide (WIDE) or long (LONG) formats – we advise using long data formats since it provides many preparation and storage benefits over wide (WIDE).

Additionally, when performing analyses on student data, additional variables must also be collected in order to identify each student’s instructor(s). The sgpData_INSTRUCTOR_NUMBER file provides an anonymous instructor-student lookup table which allows mapping student identifiers with instructors for every assessment year.

For an in-depth overview of working with longitudinal (time dependent) data with the SGP package, please consult its data analysis vignette. In general, SGP analyses require a unique student identifier, content area and instructor number, an aggregate variable to describe collective performance within each content area, scale score for assessments conducted over time and dates associated with them as variables that need to be filled in for successful analysis; any record missing any one or more will generate errors and thus cause their analyses to fail – therefore it is imperative to carefully review any sgpData_LONG file before engaging in any further SGP analyses!