Part B 1140-0091

B. Collection of Information Employing Statistical Methods

1. The customer survey is directed at customers of ATF’s arson and explosives programs and investigative assets. The specific population of customers served during a year is determined by the needs of the potential population served by ATF. Not all potential customers require the services of ATF in any given year. It is estimated that during 2003 ATF served 10,000 customers. The services provided were variable. They ranged, for example, from investigating fire and explosives scenes to responding to telephone questions on the regulations for storing fireworks. A 10% random sample (1,000) will be drawn from the population of 10,000. Sirota anticipates a response rate of 50%, or 500 responses. The sample will receive one of three surveys.

Fifty percent of the sample will receive the survey titled, “Bureau of Alcohol, Tobacco, Firearms and Explosives Arson and Explosives Customer Questionnaire.” This sample will be drawn from customers listed in the Explosives Technology Branch, Canine Operations Branch and Arson and Explosives National Repository.

Thirty-Five percent of the sample will receive the “Bureau of Alcohol, Tobacco, Firearms and Explosives Explosives Industry Customer Questionnaire.” The sample will be drawn from customers listed in the Public Safety Branch of ATF.

Fifteen percent of the sample will receive the “Bureau of Alcohol, Tobacco, Firearms and Explosives National Response Team Customer Questionnaire.” They will be drawn from the customer list of the Arson and Explosives Enforcement Branch.

The Branches with fewer customers will be oversampled to ensure a large enough number to have confidence in the results and to generalize the population as a whole. With this sample, Sirota will be able to generalize the results to the population at the 95% confidence level plus or minus 3%.

2. ATF reviewed the number of customers during 2003 for its various arson and explosives programs. Based on these numbers and oversampling programs with fewer customers, Sirota computed the sample size for each survey. The goal of the survey is to reach a representative sample of customers in order to learn of the strengths in customer service and the changes that should be made in the way ATF provides services to customers.

3. Methods to increase the response rate and minimize non-respondents include:

1. A letter from the Director of ATF to members of the sample advising that they would be receiving a questionnaire.

2. A letter from the CEO and Vice President of Sirota explaining the purpose of the survey, with guarantees of confidentially and security of data.

3. A telephone number and email address of the Vice President of Sirota to obtain information and ask questions.

4. Respondents have the alternative respond by paper and pencil or electronically.

5. Respondents have the choice of identifying themselves and/or their organization or not.

6. Respondents requiring assistance are promised follow-up.

7. Names of individuals or organizations will not appear in reports.

The sampling plan, size of questionnaire, oversampling, and the anticipated 50% response rate will permit generalization to the population with 95% confidence, plus or minus 3%. In most customer surveys conducted by Sirota, response rates of 70% to 75% are obtained.

4. The survey was pre-tested for clarity and comments on a sampling of customers before administration. The surveys have been examined by ATF and have content validity. ATF sent out sample “test” surveys, and asked respondents to comment on whether or not the survey was easy to answer and if there were questions that should be reworded or added. Comments were positive but did lead to the refinement and clarification of instructions.

5. Walter Reichman (914-696-4700, X213), Vice President, Sirota Consulting LLC or Jeff M. Saltzman. (914-696-4700, X206), CEO/Chairman, Sirota Consulting LLC.

3