cross-checking and confirmation of findings that emerge from any single method. Multiple methods are also necessary to examine the several objectives of the SBIR program, which vary in their relative tractability to quantitative and qualitative approaches. The core methodologies used in generating the data underlying the findings of this report are: surveys of firms that participated in the SBIR program; interviews with NSF SBIR officials and program managers; review of program documents and other relevant literature; analysis of program data; and case studies of firms.

3.3
METHODS USED

3.3.1
Surveys

Three surveys were sent to SBIR firms: (1) Phase II Survey; (2) Firm Survey; and (3) Phase I Survey.2

The first listed survey was focused on a sample of Phase II NSF SBIR grants. This Phase II Survey had 162 respondents. The focus of the Phase II Survey was on specific grants and their commercial outcomes.

The second survey explored the SBIR program’s influence on small businesses more generally. Responding to this survey were 137 companies identified primarily as NSF Phase II grant recipients. The survey asked questions about the background experience of each firm’s founders, the receipt of Phase I and Phase II awards, and the influence of the SBIR program on firm founding, commercial progress, and growth.

The third survey was focused on a sample of Phase I NSF SBIR grants, primarily for the purpose of identifying and learning more about those Phase I grants that did not receive a direct follow-on Phase II grant. The Phase I Survey had 248 respondents, of which 135 received Phase I grants that were not followed by a direct follow-on Phase II grant, and 113 received Phase I grants that were followed by a direct follow-on Phase II grant. Those that did receive a follow-on Phase II grant were asked if they received assistance and from whom.

The methodologies of these three surveys are described in more detail in Appendixes B and C. These descriptions reveal the initial sample sizes, adjustments to the samples, response rates, and statistics on response rates. These appendixes also provide the survey instruments as well as survey results for individual questions. The number of respondents to each question and the base numbers used to calculate percentage responses are also given. In considering these firm survey results, it is worth keeping in mind the possibility of response biases that could significantly affect the survey results. For example, it may be possible that some of the firms that could not be found have been unsuccessful and folded.

2

For the NRC Phase II Survey, 35 percent of the overall sample of 457 responded. This represents 48 percent of the awards contacted. For the NRC Phase I Survey, the response rate was 10 percent of a sample size of 2,458. See Appendixes B and C.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement