1093ss02b

4

PART B OF THE SUPPORTING STATEMENT

ENVIRONMENTAL PROTECTION AGENCY

The 2007 National Survey of Local Emergency Planning Committees (LEPCs) (Reinstatement)

1. Survey Objectives, Key Variables, and Other Preliminaries

1(a) Survey Objectives

The objectives of the survey are to: (1) track the progress of LEPCs by assessing their current activity compared to some of the data collected in a 1999 survey; and (2) probe current LEPC practices and preferences regarding several important sets of issues, including: communication with local citizens, proactive accident prevention efforts, and the effectiveness of selected OEM products and services.

1(b) Key Variables

Possible variables include: structure of LEPC; frequency of LEPC meetings; existence of emergency plan and members’ familiarity with the plan; data use and management; demographics (including type of area – rural versus urban, population size, and number of chemical facilities); communication; funding; support from federal agencies; homeland security; and assessment of strengths and areas of need. These variables are important because they will allow various cuts and sorts of the data. For example, they may help to determine correlations between LEPC area of service (both size and type) and activity level.

1(c) Statistical Approach

The 2007 LEPC survey will be administered using a universe survey or census approach, giving representatives of all LEPCs the opportunity to participate in this voluntary, one-time data collection. Both lessons learned from prior LEPC surveys and limited knowledge of the composition of the entire population of LEPCs make a census-based approach most appropriate.

In 1994, an LEPC survey was conducted using state and population-based weighted sampling. This technique was criticized by EPA stakeholders who felt that the weighting inaccurately reflected dynamics of LEPCs at local levels. Because the underlying population of LEPCs is not well known or described and LEPC service areas vary tremendously in terms of industry, geographic size, and population, selecting a sampling frame that accurately represents the entire population of the nation’s LEPCs would be very difficult.

In response to criticism following the 1994 survey and in consideration of the challenges of selecting a sampling frame that accurately reflects the target population, the 2007 LEPC survey (like the 1999 LEPC survey) will be administered to all known LEPCs. Thus, unless warranted (i.e., all 3,500 LEPCs respond), statistical inferences will not be used to make generalizations about the entire population of LEPCs. Reports derived from this information collection will focus only on those individuals who voluntarily respond to the survey.

EPA will engage SRA International, Inc. (3434 Washington Blvd., Arlington, VA 22201) to help design, administer, and provide technical assistance and support for the survey. Furthermore, this contractor will analyze the data and prepare the draft report.

1(d) Feasibility

While recognizing that respondents may encounter certain difficulties (e.g., little or no time to take the survey), we believe that it is feasible to efficiently conduct a national survey of LEPCs because the survey is designed to be as easy as possible to complete, drawing on knowledge that the respondents should be able to accurately recall without the need for research. Another possible challenge for LEPCs may be limited Internet access. OEM does not anticipate many LEPCs to be without Internet access, considering communication and outreach is integral to emergency planning. Neither of these possible obstacles were evident in EPA’s two previous surveys of LEPCs. Sufficient funds are available to complete the survey as designed and survey results will serve OEM’s decision-making needs.

2. Survey Design

2(a) Target Population and Coverage

The survey will target 3,500 LEPCs nationwide. Survey coverage will focus on one representative in a leadership position from each LEPC.

2(b) Sample Design

One representative in a leadership position from each LEPC (approximately 3,500) will be targeted. E-mail addresses for these individuals will be obtained from EPA’s LEPC database, which is in the public domain and was recently updated at: http://yosemite.epa.gov/oswer/lepcdb.nsf/HomePage?openForm. Due to the nature of a universe survey and the information contained in the LEPC database, respondents will be invited to take the survey on a nonrandom basis. While the sample design is nonrandom, the leaders who will take this survey best represent the LEPCs on which they serve.

2(c) Precision Requirements

The potential sample size is 3,500 (one representative from each LEPC). As the survey is completely voluntary and EPA seeks the broadest inputs of stakeholders, no stratification of respondents is planned. The purpose of this survey is to take a "pulse check" of LEPCs who choose to complete the survey. Steps have been taken to limit response errors and misinterpretation of the questions as well as to minimize potential errors of collection, response, and processing of data. Regarding nonresponse, LEPCs who do not respond are likely inactive and have little vested interest in the survey results. For this reason, it is unlikely that non-response will have a significant effect on the value of the survey as a decision-making tool.

2(d) Questionnaire Design

Two previous surveys of LEPCs were administered in 1994 and 1999. Several questions from both the 1994 and 1999 instruments have been incorporated into the 2007 National Survey of LEPCs. Questions were written so that they would be easy to read and understand and include common terminology. The wording is simple, the answer categories precise, the layout attractive, and the questions are organized into specific categories and follow a logical pattern. A brief description of each question is given under the section 4(b)(i) of this Supporting Statement’s Part A.

Furthermore, the survey is designed to be administered over the Web. For this reason, the questionnaire is formatted in a way that is user-friendly. The layout is clear and the skip patterns are logical and function appropriately. The survey clearly articulates that the information provided by users is confidential and that their participation is entirely voluntary. A prototype of the survey instrument is attached to this ICR.

3. Pretests and Pilot Tests

EPA will request six people to test the survey. The six people will be selected from the following stakeholder groups: EPA headquarters, EPA Regional Office, SERC representative, two LEPC representatives, and a representative from industry.

4. Collection Methods and Follow-up

4(a) Collection Methods

The survey will be administered electronically through a web-based application. This collection method was selected because of its reliability, usability, precision, and cost effectiveness. EPA will e-mail an invitation to the pre-identified LEPC representatives, directing them to the Web site survey application and providing potential respondents with survey log-in information (User ID and password).

4(b) Survey Response and Follow-up

It is anticipated that this survey will have a response rate of approximately 50%. This is based on past experience (specifically the 1999 nationwide survey of LEPCs). To increase the response rate, EPA will prepare a follow-up e-mail to remind respondents that the deadline for completing the survey is approaching. This second e-mail will only be sent to LEPC contacts who have not yet taken the survey. Furthermore, EPA will post survey results on its Web site as an incentive for LEPCs to participate in the survey.

5. Analyzing and Reporting Survey Results

5(a) Data Preparation

The web-based application will automatically collect the data into predetermined

fields for analysis. SRA International, Inc. will support the preparation of the data and will conduct data quality checks to ensure data has been appropriately categorized and sorted.

5(b) Analysis

The web-based application will automatically collect the data into predetermined fields for analysis. Excel spreadsheets will be created that provide breakouts of data by the response categories. A summary report will be prepared with the results of the survey and will include appropriate graphs and charts with key findings and highlights.

5(c) Reporting Results

Agency personnel will not have direct access to the database. The database will be housed on the contractor’s external server and only aggregate data will be reported. Individual records will not be attributed to individual respondents. Instead, a summary report will be prepared of overall key findings as well as cross-tabs by demographic.