Among the many facets of the SGP package is its ability to perform analyses on sgpData, an anonymized panel data set. The sgpData table contains five columns, each representing a student assessment score for a specific year. The table also shows NAs for missing data.
The first column of sgpData must contain a unique student identifier. The next column must provide the grade level and/or time associated with assessment occurrences. The next set of columns must also provide numeric scores for each assessment occurrence. The sgpData table is the basis for most SGP analyses, as the data represents a scale score for each year. However, there are a number of other variables that are used for demographic purposes. These include LAST_NAME, FIRST_NAME, GRADE, ACHIEVEMENT_LEVEL, SCALE_SCORE, and CONTENT_AREA.
There are also many different ways to perform SGP analyses. For example, the summarizeSGP function can be used to generate student aggregates using all of the demographic variables that are provided in the sgpData table. This function uses the SCALE_SCORE variable to generate student growth projections. It is also possible to use the summarizeSGP function to generate student growth plots at the individual level. However, most SGP analyses are best done using the LONG format. It has several benefits over the WIDE format, including the ability to spread time dependent data across multiple rows. It also has some advantages in terms of storage, as the data can be more accurately represented as a time series.
The sgpData table also includes a column for each year that contains the missing value (NA) for the assessment score. This is useful for analyses involving missing data. As the missing value may not be available for all years, it can be a useful tool to examine trends in assessment data.
Finally, the SGP package also provides a comprehensive documentation for sgpData. This includes an exemplar data set, as well as numerous examples of how to use the data. The documentation includes a detailed explanation of each of the functions used to run SGP analyses. The package also assumes that the state-specific meta-data is available.
Finally, the SGP package offers an updated version of its sgpData function, called updateSGP, which can be used to wrap all of the steps required for SGP analyses into a single function call. This function makes the source code easier to understand by wrapping six steps into a single function. However, this function can cause a site error if a lantaran is used.
The sgpData table is also useful for running analyses that involve growth plots. This function can be used to generate student growth plots at the individual and school level. It also allows for analyses at the school and district level. The package also includes a number of exemplar longitudinal data sets. Each exemplar data set contains a data set that reflects the most important features of the sgpData table. However, these data sets may not be available in all states.