The Data Analysis System (DAS) is a software application that allows you to produce tables and correlation matrices from NCES data sets. There is a separate DAS for each data set, but all have a consistent interface and command structure. Each DAS runs in two modes-Tables and Correlations-and is available in both a Windows- and web-based format. Using the DAS Online, you can create programming instruction files, run tables and correlation matrices, and view output using a web browser.

Tables

In the Tables mode, the DAS will create a table of estimates, corresponding standard errors that are calculated by taking into account the complex sampling designs used in NCES surveys, and weighted sample sizes for the estimates. Please note that the weight variables for the Online and Windows versions differ. When you import a TPF/CPF that contains a weight variable from the Windows version, the DAS Online will automatically replace the Windows weight with an appropriate Online weight.

Correlations

In the Correlations mode, the DAS will produce a correlation matrix that can be used for linear regression models. The design effects (DEFTs) for each variable are included in the output. Since statistical procedures generally compute regression coefficients based on a simple random sample assumptions, the standard errors must be adjusted with the design effect to take into account the stratified sampling method used in NCES surveys.

There are two utilities: the Batch Processor and the Weight Convertor.

Batch Processor

Using the Batch Processor feature, you can import TPFs and CPFs from multiple data sets and run them in one job. In order to run TPFs and CPFs containing old weight variables from DAS for Windows, you must first use the weight convertor.

Weight Converter

The Weight Converter will replace Windows weight variables with Online weight variables. You can submit one TPF/CPF indvidually or many compressed into a zip file. Once the weight variables are changed, you can submit the new zip file to the Batch Processor.