ED Supporting Statemen Part B

Supporting Statement: Part B

Collections of Information Employing Statistical Methods

1. Sampling or other respondent selection methods

The respondent universe for this survey consists of all individuals who sought assistance from SBA and its resource partners during a fiscal quarter (3 months) of a previous fiscal year selected randomly. Annually, SBA through its resource partners provides face-to-face counseling to over annually which equates to a month. On average, this works out to be 83,333 individuals per month.

The initial survey sample will consist of a 24 percent sample randomly drawn from clients served during a fiscal quarter (3 months) of a previous fiscal year. At present, the SBA plans to implement the survey annually. A stratified random sample will be implemented to select a nationally representative, self-weighting sample. Individuals will be stratified by the program they contacted for assistance (SBDC, SCORE and WBC), and the month and year they contacted the program. At the first stage of the selection process, two non-contiguous months will be randomly selected. The sample will be selected to proportionately reflect the clients served by each program as a percentage of the total clients served by all of the programs. Assuming that 54 percent of the individuals are SBDC clients, 37 percent SCORE clients, and 9 percent WBC clients, the sample will look as follows:

For a Fiscal Quarter

SBDC Clients: 12222

SCORE Clients: 11300

WBC 1,496

Total 25048

Parsing of economic impact

After discussions with the Chief Economist of the SBA and the Senior Research Fellow at the NFIB Foundation, SBA will attempt to measure impact against two control groups. For the Nascent entrepreneurs, we will compare results against the PSED sample from the Kauffman Foundation. To do this, we need to add two questions to the proposed surveys – Age of respondent and whether they had previously owned a business.

For the second control group, comparing early start-up and in-business clients we will either use a sample drawn from Dun & Bradstreet data file matching against firm size and industry or draw a sample of SBA loan recipients. For the loan recipients, we will screen out any respondent who indicates they also used ED resources [SCORE, SBDC and WBC] during the time period.

Statistical Analyses we will employ on data will be Frequencies, T-tests and Regressions once we have at least two data points - Year 1 and Year 2, economic data. Given we are comparing matched pairs; T-Tests and possibly chi-square analyses are the most logical.

Non-response bias

An extensive telephone follow-up will be initiated.

2. Procedures for the collection of information

a. See attached research plan. Further we believe a 25 percent response rate with extensive non-responsive activities should provide a representative sample for statistical analysis.

3. Methods to maximize response rates and to deal with issues of non-response

See attached research plan

4. Tests of procedures or methods to be undertaken

Since the inception of the survey, we have changed the method of data collection to be more efficient by scrubbing the data before mailing to validate addresses. We are implementing a pilot using a sample of SBDC clients to determine the feasibility of online surveys.

Individuals consulted:

The following individuals were consulted on the statistical methods to be used in this information collection:

Dr. K. Mark Weaver

Thomas H. Daigre Endowed Chair of Business Administration
Louisiana State University
3152 B CEBA Building
Baton Rouge, LA 70803
Principal Investigator National SBI Studies

Dr. Chad Moutray

Chief Economist SBA

Office of Advocacy

US Small Business Administration

Washington, DC

William Dennis

Senior Research Fellow

NFIB Foundation

Washington, DC

Dr. Patricia Greene

Provost, Babson College

Wellesley, MA

Former Ewing Marion Kauffman/Missouri Chair in Entrepreneurial Leadership

Henry W. Bloch School of Business and Public Administration

Principal Investigator “Diana “Project National Study on Women Entrepreneurs

Dr. Frank Hoy

Former Dean University of Texas at El Paso

College of Business Administration

El Paso, TX

1^st Principal Investigator “ASBDC Studies on Economic Impact

Dr. Bruce Kirchhoff

Distinguished University Professor

School of Industrial Management

New Jersey Institute of Technology

University Heights

Newark, NJ

Former Chief Economist SBA

Research Plan

Impact Studies

Office of Entrepreneurial Development

1. Research Objective

1. Measure outcome of services and assistance provided multiple small business market segments [i.e., Nascent, Start-Ups and In Business] through multiple SBA ED resources [i.e., SBDC, SCORE and WBC] regarding three primary measures [i.e., attitudinal, improvement in management/marketing skills and business growth] for the discrete market segments.

2. Research Questions

   1. There is a positive relationship between the services/assistance provided by [i.e., SBDC, SCORE and WBC] and the client’s perception of the service/assistance received.

   2. There is a positive relationship between the services/assistance provided by [i.e., SBDC, SCORE and WBC] and the client’s decision to start a business.

   3. There is a positive relationship between the services/assistance provided by [i.e., SBDC, SCORE and WBC] and the client’s decision to implement and/or change management marketing practices in their business.

   4. There is a positive relationship between the services/assistance provided by [i.e., SBDC, SCORE and WBC] and the client’s business having a positive financial impact [jobs created and retained, increase in gross sales].

3. Define Sampling Frame

1. By Respondent

   1. Nascent [Kauffman Definition those whose intention to start a business is greater than the general population]

2. Start-up [Those individuals who have expressed a desire to start a business in the next one to nine months]

3. In Business [Those businesses classified as small by the SBA]

2. By ED Resource

   1. SBDC

   2. SCORE

   3. WBC

4. Sampling methodology

1. Stratified Random Sampling

1. Resources will select clients served within a ninety day time period and using random table mail out surveys to designated stratification.

2. Contractor will provide a table of random numbers to select survey clients based on respondent characteristics and annual percent of clients served.

3. All surveys will be coded to ensure that addresses and contact information remain with resources and ensure follow-up for non-response.

4. After initial survey, those clients in business or indicating strong preference to start business will be surveyed again in 12 months using the attached questionnaire.

5. All responses mailed back to resource to increase likelihood of response.

6. All completed responses sent to contractor for analysis.

5. Non-response Bias for initial/baseline questionnaire

1. After 10 days from time of mailing by contractor, a follow-up package will be mailed to non-respondents.

2. After mailing of package, telephone and/or email follow-up will occur on all non-respondents.

6. Non-response Bias for follow-up questionnaire [Questionnaire B]

1. After mailing of package, telephone and/or email follow-up will occur on all non-respondents.

7. Time Frame

   1. Obtain OMB clearance 4^th quarter FY 2009

   2. Write RFP to solicit contractor 4^th quarter FY 2009

   3. Begin collecting baseline data 1^st quarter FY 2010

      1. Baseline data from 4^th quarter FY 2008

   4. Begin data analysis of baseline 2^nd quarter FY 2010

   5. Write draft report 3^rd quarter FY 2010

   6. Present Final report 4^th quarter FY 2010

   7. Begin collecting follow-up data 1^st quarter FY 2011

   8. Collect Follow-up data from 4^th quarter FY 2010

   9. Again collect baseline data 1^st quarter FY 2011

      1. Baseline data from 4^th quarter FY 2009

10. Begin data analysis of baseline and

follow-up data 2^nd quarter FY 20 11

11. Write draft reports 3^rd quarter FY 2011

12. Present Final reports 4^th quarter FY 2011