The goal of data sgp is to create and make available to researchers the multi-proxy geochemical data needed for analysis of shale deposits over the Neoproterozoic through Paleozoic epochs. This work is being conducted within the context of a research consortia (rather than a full community database such as Genbank or EarthChem). Although the resulting data may be eventually integrated with these community databases, the original motivation for this work is to provide the tools and methods to address specific scientific questions.

This is a large amount of data, but not really ‘big data’ in the sense that it is not too big for standard statistical software applications or computational environments. In fact, if compared to the data required for a large cosmological simulation or an analysis of social network interactions it is quite small. However, it does represent a substantial leap in scale from previous sedimentary geochemical research and will require a significant investment of both computing and human resources to make the most of it.

The SGP package installs an exemplar longitudinal (time dependent) data set sgpData_LONG and INSTRUCTOR-STUDENT lookup file sgpData_INSTRUCTOR_NUMBER to assist in the creation of this kind of data. Typically, longitudinal data is stored in WIDE format with each row representing a student and each column an assessment occurrence. The sgpData_LONG data set models this format and the lower level studentGrowthPercentiles and studentGrowthProjections functions can be used with this or any other longitudinal data.

For operational SGP analyses, a student’s performance is usually evaluated against the performance of his/her peers. To do this, the students are grouped into groups called ‘classes’ and each class is given a’score’ that represents its relative standing in the group. These scores are then used to calculate a student’s overall average score. This score is then compared against the state standard to determine a students grade.

In order to calculate these scores, the NJDOE links district course roster submission data with relevant mSGP teacher evaluation ratings and tabulates these with the teachers’ teacher practice score and SGO score. The mSGPs are then converted to a 1-4 score using the Median Student Growth Percentile Conversion Chart. This process is not simple and requires careful verification of the data by both districts and individual teachers. However, it is a necessary step in ensuring that the mSGPs are accurate and reflect actual classroom performance. This is an important element in the quality of educational decisions made by districts and the NJDOE. It is also a key part of the transparency of SGP results. This helps to maintain a high level of trust in the system. The SGP team is working hard to improve this process and encourages users to contact them with their suggestions.