Business Structure
The data for this survey report were collected for the NFIB Research Foundation by the executive interviewing group of The Gallup Organization. The interviews for this edition of the Poll were conducted in December 2004 from a sample of small employers. “Small employer” was defined for purposes of this survey as a business owner employing no less than one individual in addition to the owner(s) and no more than 249.
The sampling frame used for the survey was drawn at the Foundation’s direction from the files of the Dun & Bradstreet Corporation, an imperfect file but the best currently available for public use. A random stratified sample design is typically employed to compensate for the highly skewed distribution of small-business owners by employee size of firm (Table A1). Almost 60 percent of employers in the United States employ just one to four people meaning that a random sample would yield comparatively few larger small employers to interview. Since size within the small-business population is often an important differentiating variable, it is important that an adequate number of interviews be conducted among those employing more than 10 people. The interview quotas established to achieve these added interviews from larger, small-business owners are arbitrary but adequate to allow independent examination of the 10-19 and 20-249 employee size classes as well as the 1-9 employee size group.
Table A1
Sample Composition under Varying Scenarios
|
Expected from |
Stratified Random Sample |
Obtained from Stratified |
Employee Size of Firm |
Interviews |
Percent Distri- |
Interview |
Percent Distri- |
Completed Interviews |
Percent Distri- |
1 - 9 |
593 |
79 |
350 |
47 |
353 |
47 |
All Firms |
750 |
100 |
750 |
101 |
752 |
101 |