Supporting Statement Part B for MPS-Revised

2

SUPPORTING STATEMENT FOR THE MERIT PRINCIPLES SURVEY

Part B. Collection of Information Employing Statistical Methods

1. Description of the potential respondent universe and any sampling selection methods to be used.

The population to be studied in the Merit Principles Survey is comprised of the full-time, permanent, non-seasonal civilian workforce of 1.9 million employees in the executive branch agencies, excluding the Postal Service. To ensure representation across the diverse population of Federal employees, we will utilize random sampling from the Civilian Personnel Data File (CPDF) to select employees in each participating Federal agency (or agency component for large agencies) stratified by supervisory status. Our goal is to obtain in each agency or component at least 400 responses from supervisors and managers and 400 responses from employees. In small agencies with fewer than 400 employees and supervisors, we plan to include all employees and supervisors in our sample.

We expect an average return rate of 50% and therefore we will draw a sample of approximately twice the number of returns expected for each strata (800 -1,000 employees and 800-1,000 supervisors/managers) for each agency or component. Twenty-eight agencies will participate in the survey. For large agencies, subcomponents will be separately sampled and reported. The total number of persons included in the sample will be about 80,000 with an estimated 40,000 respondents.

2. Description of procedures for the collection of information.

A probability sample will be drawn in which each employee in the target population has a known, non-zero probability of selection so that results can be generalized from the survey respondents to the survey population. The sampling frame will be the lists of employees from each participating agency grouped into employees and supervisors. From each of these lists, 800 – 1,000 employees and 800 – 1,000 supervisors will be randomly selected.

3. Description of methods to maximize response rates.

To maximize return rates, we will provide participating agencies with posters, flyers, the content for e-mail messages, and other communication materials to distribute to their employees before the survey is distributed. These materials are intended to build the interest and enthusiasm among employees required for a high response rate. We will be working with a contact person in each agency whom we will coach to answer employees’ preliminary questions.

We will individually invite the prospective respondents to participate in the survey with an email message or cover letter which will be signed by an MSPB official. This message will explain how the employees were selected and how the confidentiality of their responses will be maintained, while stressing the importance of the research and their participation to ensure representation of all perspectives. We will also request that the agency leader express support for completing the survey so employees will be assured they can complete the survey during work time.

We will follow up with weekly email messages as a reminder to non-respondents. These messages will contain everything needed to access the survey (i.e., user ID, password, and hypertext link to the secure survey site).

We plan to send both the initial invitation emails and reminders from an mspb.gov address to provide additional credibility to our survey and help Federal employees distinguish it from requests that they may receive from other sources or spam.

Respondents who encounter difficulties accessing or completing the survey will be able to contact a “help desk” for assistance. Respondents will also have the option of requesting a paper copy of the survey if they prefer.

We plan to examine the response rate for each of the stratified groups each week, and when needed, conduct additional follow up efforts to ensure greater participation.

Finally, in the design of the survey instrument, we pilot tested the questions to ensure that the total length is reasonable and that the questions are clear since extremely lengthy surveys and/or difficult questions may discourage respondents from completing the survey. We also created a question format that is user-friendly and efficient.

4. Description of test procedures.

We will conduct a preliminary field test of the survey prior to administration by having a few members of our own staff complete the survey. After revising the survey as needed and placing it online, we will ask additional staff from the Office of Policy and Evaluation and other offices within MSPB to complete the survey online to help ensure a user-friendly navigation through the survey. We also plan to ask field testers in each agency to try out the survey link to test its functionality before we send e-mail invitations to our sample of participants.

5. Name and telephone number of individuals consulted.

Persons consulted on the statistical design:

John Crum, MSPB, (202) 653-6772, ext. 1344

Persons who will collect and/or analyze the information:

Anne Marrelli, MSPB, (202) 653-6772, ext. 1341

Cynthia Ferentinos, MSPB, (202) 653-6772, ext. 1334

John Ford, MSPB, (202) 653-6772, ext. 1104

James Tsugawa, MSPB, (202) 653-6772, ext. 1338