To maintain high quality in the dissemination of data from national data bases, our approach begins with the development of quality assurance tests for any data files to be used in an analysis. For the SCS data collection, this includes checks for internal consistency of the data, comparisons of estimates with previous years of the collection, and alignment of variable values with data definitions provided in any user documentation. Our approach to quality assurance for the tables, figures and reports we produce for NCES includes three levels of checks.
First, all estimates are independently generated twice – ensuring that any assumptions made and statistical programs developed to generate the information are verified.
Second, these estimates are transformed from output produced by statistical programs such as SAS or SPSS into formatted tables for publication using macros to avoid errors that might arise from retyping figures, and which concurrently apply NCES statistical standards such as checking the ratio of standard errors against estimates and replacing table entries with warning symbols where estimates do not meet statistically valid thresholds.
Third, all materials developed by our analysts are reviewed by Synergy’s experienced communications staff to ensure formatting, labeling and explanatory text meet NCES publication standards.
Synergy maintains quality standards throughout NCES’ publication review process, responding to comments, performing additional analysis when requested, and maintaining careful version control as edits are made.