The Business R&D and Innovation Survey - Lessons Learned from the Redesign of the Survey of Industrial Research and Development

Sep 19, 2026

Upload by Himbo

The Business R&D and Innovation Survey paper explains how the U.S. Census Bureau and NSF redesigned the Survey of Industrial Research and Development (SIRD) to produce higher-quality, relevant R&D data for the business sector. It contrasts earlier survey limits with the post-1953 SIRD’s original financial focus, then details the redesign rationale and the five-phase process: defining needs, identifying data sources, evaluating methods, implementing and testing new content via cognitive testing, and evaluating pilot results through analysis and post-survey interviews.

LanguageEnglish

CategoryPresentations

FormatPDF

Pages11

Reading Time4 min

AccessFree to download

Background

Five Phase Redesign

Phase 1: Define data/information needs and priorities

Phase 2: Identify data sources and availability

Phase 3: Evaluate current operations and methodology

Phase 4: Implementation and testing of new content and methods

Phase 5: Evaluation of the pilot results

Changes to Survey Content

Adopting a definition of R&D based on accounting standards

The Business R&D and Innovation Survey: Lessons Learned from the Redesign of the Survey of Industrial Research and Development Richard S. Hough U.S. Census Bureau Brandon Shackelford Twin Ravens Consulting John Jankowski, Ray Wolfe National Science Foundation Federal Committee on Statistical Methodology Research Conference January 10-12, 2012 Background Before 1953, federal statistics could tell you how many widget-making plants there were in the U.S. and what their sales of widgets were. What they couldn‘t tell you was the amount of research and development companies were performing to―build a better widget‖—or to develop any other new or improved good or service. However, since 1953 the Survey of Industrial Research and Development (SIRD) has collected information, primarily financial in nature, on the systematic work companies were undertaking with the goal of discovering new knowledge or using existing knowledge to develop new or improved goods and services. The SIRD was sponsored by the National Science Foundation‘s (NSF) National Center for Science and Engineering Statistics and was conducted by the U.S. Census Bureau in accord with an interagency agreement through 2007. Over the past 50+ years, NSF and the U.S. Census Bureau added and deleted content from the SIRD based on input from companies responding to the survey as well as users of the data, but the fundamental concepts and structure of the survey were largely untouched. More recently, prompted by recommendations from the 2005 Committee on National Statistics (CNSTAT) Report, Measuring Research and Development Expenditures in the U.S. Economy, NSF and the U.S. Census Bureau concluded a full-scale redesign of the SIRD. The goal of the redesign was to produce high quality, relevant data on R&D in the business sector that took into account the changing reality of R&D and innovation writ large. This paper will discuss the redesign project and some of the most significant changes that were implemented in the 2008 pilot of the redesigned survey, titled the Business R&D and Innovation Survey (BRDIS) . Five Phase Redesign A full-scale survey redesign involves a range of required tasks. In order to simplify planning and management, the SIRD interagency redesign team organized these tasks into five phases: Phase 1: Define data/information needs and priorities. To accomplish the goals of this phase the interagency team conducted two data user needs workshops (federal and non-federal data users); held three panel meetings of R&D managers and executives from a variety of U.S. firms; conducted a series of innovation workshops; reviewed existing literature that referenced the survey data; and evaluated the data needs of NSF, the Bureau of Economic Analysis (BEA), and the U.S. Census Bureau. Phase 2: Identify data sources and availability. The team conducted record keeping studies, respondent debriefings, cognitive interviews, response studies of the current SIRD and used supplemental surveys to gather information from the current set of respondents. The team also consulted with a panel of business experts to better understand what information firms can provide and are willing to provide. Phase 3: Evaluate current operations and methodology. During this phase, the team examined the industry classification, imputation and editing methodologies implemented in the SIRD and identified areas where new methods and technologies would improve the results. Phase 4: Implementation and testing of new content and methods. This phase synthesized the information gathered during the first three phases to implement the new Business R&D and Innovation Survey. A critical component of this phase is the cognitive testing of proposed survey content with potential respondents. Phase 5: Evaluation of the pilot results. This is the post collection phase where the reported data is analyzed for accuracy and consistency with desired concepts. The critical compon

FAQs

Why was the SIRD redesigned into the Business R&D and Innovation Survey (BRDIS)?

The redesign aimed to generate higher-quality, more relevant R&D data for the business sector while accounting for changing realities of R&D and innovation.

What are the five phases of the full-scale survey redesign?

The process covers defining needs and priorities, identifying data sources and availability, evaluating current operations and methodology, implementing and testing new content and methods, and evaluating pilot results.

How did the BRDIS define R&D for data collection?

The survey adopted an R&D definition based on accounting standards, starting from the companies’ accounting understanding and deriving the performance-based concept from that information.

Sep 19, 2026

Upload by Himbo