revised supporting statement SPDO B

B. Statistical Methods

1. Universe and Respondent Selection

The Survey of Public Defender Offices is a census of all state- and county– funded public defender offices across the county. The principal analyses of the data collected will be at the national, state, and county levels. . It is estimated that there are approximately XXXXX 1,400 public defender offices to be included in the censussurvey effort. The principal analyses of the data collected will be at the national, state, and county levels.

As is standard for BJS surveys, Tthe response rate expected is 950 percent or more. In 1999, the county survey yielded a response rate of 94 percent and the program survey a response rate of 95 percent. Other, more recent BJS data collection efforts, including Survey of Prosecutors, Census of Medical Examiners and Coroners Offices, and Census of Publicly Funded Forensic Crime Laboratories, have successfully met and/or exceeded this mark.

2. Procedures for Collection Information

The National Survey of Indigent Defense SystemsPublic Defender Offices will is a nationwide census of all state-and locally- funded public defender offices across the country. It is estimated that there are approximately 1,400 such offices. This estimate is based primarily on the National Legal Aid and Defenders Association (NLADA) reliable and fairly exhaustive directory and will be validated through state and local court contacts. The respondent contact list will be updated prior to data collection in order to maintain a current list of public defender offices

There is good justification for surveying the full universe. The data collection represents the first complete census of public defender offices and, as such, could be used as a roster for future sample surveys. State- and locally- funded public defender offices handle a substantial portion of indigent defense cases but differ widely from jurisdiction to jurisdiction in terms of organization, funding sources, and types and characteristics of cases handled. The resources (e.g., budget and staffing) available to smaller public defender offices, for example, can be very different than those in larger offices and most likely affect policy decisions. It would be beneficial to the indigent defense community to report data on the policies and practices of the wide range of public defender offices and to display the data on an interactive web site that includes all offices throughout the states. A sample survey would not allow data to be presented for each office.

Different data collection strategies may be necessary to effectively gather required information from the variety of office types. Therefore, respondents will be provided with two different delivery options, a paper instrument that can be mailed or faxed back to the data collection agent and a web-based version of the questionnaire that can be use probability sampling to ensure that the variance estimates can be developed for survey estimates derived from the data. County probabilities will employ 1997 population counts as measures of size. Independent samples will be drawn for each state. The most 100 populous counties will be included with certainty. Roughly equal numbers of sample counties will be drawn, as in the 1982 design, ranging from 15 to 18. While the state is the primary stratification variable by virtue of independent samples drawn within states, additional stratification will include: population size, urban/rural county status, and geographic indigent defense system centralization for those states which have organized their indigent defense systems by county based on geographic groupings (e.g., judicial districts.)filled out on-lines. Standard practices follow-up practices, such as sending reminder post-cards and additional copies of the survey, will be utilized for non-respondents. The most recalcitrant offices will be allowed to provide information via phone interview.

The reference year for the county and program questionnaires is fiscal calendar year 19972006. County and program questionnaires will be mailed to rRespondents, who will be asked to complete the survey, either manually or electronically and return it within 30 days2 weeks.

3. Methods to Maximize Response

To maximize response rates BJS will employ the best practices for survey nonresponse follow-up. We will do everything possible to maximize response, including the use of letters, telephone interviews and fax. These documents will be developed through collaboration between BJS and the data collection agent.

Initially, a follow-up post card will be sent to thank respondents who have returned the questionnaires and to serve as a reminder for those who have not yet completed the survey. Telephone prompting will be begin approximately ten days after the postcards have been sent to confirm receipt of the survey, explain the purposes of the study, answer any questions respondents may have, avert potential refusals and establish an expected date for the return of the questionnaire. Telephone interviews with the remaining non-respondents will begin about two weeks after the postcard mailout. This strategy has been successfully employed by the National Opinion Research Center in a number of studies with respondents in institutional settings to yield a response rate of over 95 percent.

4. Pre-test

A pre-test of the questionnaire will be conducted in in XX6 public defender offices. This pre-test will help BJS evaluate the effectiveness of the survey forms and the ability of respondents to answer the questions. It will also help BJS modify certain aspects of the survey forms to make it more efficient, which in turn will lessen the burden on respondents.

5. Contacts for Statistical Aspects and Data Collection

Lead agency: Bureau of Justice Statistics

U.S. Department of Justice

Thomas Cohen, Ph.D., Statistician

Law Enforcement, Adjudication and Federal Statistics Unit

Bureau of Justice Statistics

810 Seventh Street, N.W.

Washington, D.C. 20531

(202) 514-8344

Lynn Langton, M.A., Statistician

Law Enforcement, Adjudication and Federal Statistics Unit

Bureau of Justice Statistics

810 Seventh Street, N.W.

Washington, D.C. 20531

(202) 353-3328

Steven K. Smith, Ph.D., Chief

Law Enforcement, Adjudication and Federal Statistics Unit

Bureau of Justice Statistics

810 Seventh Street, N.W.

Washington, D.C. 20531

(202) 616-3485

Contractor: The Spangenberg Group

Robert Spangenberg, President

The Spangenberg Group

1001 Watertown Street

West Newton, MA 02165

(617) 969-3820

C. Attachments

1. OMB form 83-I Paperwork Reduction Act Submission

2. Certification Statement

3. Draft of Survey of Public Defender Offices questionnaire

4. Law or authority mandating the information collection

5. 60-day ICR notice that will be published in the Federal Register

6.. 30-day ICR notice that will be published in the Federal Register

1. Copy of 60 day Federal Register Notice.

2. Copies of the NSIDS-2 county and program survey forms.

3. Copy of the regulatory authority (42 U.S.C. 3732)