Papers
|
CAI Survey Development and Documentation System (SuDDS) Richard L. Downs, Jr. U.S. Census Bureau |
|
Abstract
SuDDS is a database application that facilitates computer-assisted interviewing (CAI) specification, documentation, testing, problem tracking, and processing. Its purpose is to create a uniform CAI specification methodology that allows for input from various sources (an IDE, IDOCS, word processing templates, other specification database formats), places various verification/analysis tools at analysts disposal, and produces uniform outputs for internal and external customers. Examples of tools include a suite of tools to check for irregularities in the specification. Examples of outputs include a suite of Blaise instrument specifications, CASES specifications, change requests, and an items booklet.
Another Simple SAS/AF Program to
Manage SAS/CONNECT Sessions
Richard L. Downs, Jr.
U.S. Census Bureau
Abstract
This paper highlights a simple SAS/AF program that assists with several basic SAS/CONNECT tasks. A user only needs to know the computer's name to sign on or sign off using a point-and-click approach. The program also handles defining libraries on other computers and reassigning these libraries the next time the user connects. Additionally, the program uploads/downloads SAS data sets and catalogs and submits SAS programs on the other computer.
Bio
Richard L. Downs, Jr. has a B.S. in Information Systems Management from the University of Maryland, Baltimore County and a M.S. in Computer Systems management from the University of Maryland. He has worked at the U.S. Bureau of the Census for 13 years as an applications and systems programmer with 8 years programming experience in SAS. He has presented at SUGI, NESUG and SESUG.
|
A Simple Framework for Sequentially Processing
Hierarchical Data Sets for Large Surveys
Richard L. Downs, Jr. and Pura A. Peréz
U.S. Census Bureau
Abstract
This paper explains a simple framework for applying traditional sequential processes (reformat, edits, imputation, etc.) to a hierarchical group of data sets. The framework merges the data sets using a series of layered data step views. It then shows how users can write data step code that processes the merged data sets and outputs updated versions of one or more of those data sets. The data step code uses a macro to help dynamically create a keep option for each output data set. The data step then uses a combination of checking for last. variables and missing values to determine the proper time to output to each data set. This framework largely isolates the complexity of relating the data sets from the complexity of the actual process, it maintains the data set hierarchy, and eliminates the need for any post-process processing (all output data sets are complete at the end of the process data step).
Bios
Richard L. Downs, Jr. has a B.S. in Information Systems Management from the University of Maryland, Baltimore County and a M.S. in Computer Systems management from the University of Maryland. He has worked at the U.S. Bureau of the Census for 10 years as an applications and systems programmer with 5 years programming experience in SAS.
Pura A. Peréz has a B.S. in Mathematics and Computer Science from the Catholic University of Puerto Rico and a M.S. in Computer Systems Management with a concentration in Database Systems and Security from the University of Maryland. She has worked for the past 7 years at the U.S. Bureau of the Census as an applications programmer with 3 years experience programming in SAS.