Att_SupprtStateB

B. Statistical Methods

B1. Sampling

None of the data collection included in this study involves sampling.

B2. Procedures for the Collection of Information

a. Sample Selection

No sampling will be done as part of this study.

b. Estimation Procedures and Degree of Accuracy

Because there will be no sampling, estimation procedures will not be employed.

c. Data Collection Procedures

The two questionnaires will be placed on Development Associates’ Web site. To initiate the survey, e-mails will be sent to each State VR Director’s and SRC Chairperson’s individual e-mail address. Each e-mail will contain the URL address of the relevant questionnaire as well as a password that is unique to each VR agency and SRC. In the case of VR agencies, there will be two passwords, one for the State Director and one for all other respondents in the agency. The passwords provided to State Directors and SRC Chairpersons will be used to submit the final questionnaire responses to Development Associates. If requested, we will also provide versions of the forms in alternative formats.

B3. Procedures to Maximize the Response Rate

The goal for this needs assessment will be a 100 percent response rate from both the State VR agencies and the SRCs. We believe that the flexible nature of a Web-based survey will facilitate responses. We also believe that because the results will be used to design a technical assistance network for State VR agencies and SRCs, respondents will be highly motivated to respond.

Our approach to maximizing response rate will include e-mail reminders, follow-up telephone calls, and the presence of a toll-free 800 number which respondents can call with questions or to request assistance. Using comparable follow-up methods on a previous study for RSA, we received a 100 percent response rate on a survey of more than 50 American Indian VR projects.

B4. Pilot Tests of Data Collection Procedures and Instruments

Web-based versions of the instruments were pilot-tested with three State VR agencies (Georgia combined, Missouri general, and New Mexico blind) and three SRCs (New Jersey general, Missouri general, and New Mexico blind).

Based on the results of these pilot tests, we have made two important changes to the data collection instruments:

• The wording of the scale for the first item on all of the forms has been changed from “Little importance……Great importance” to “Little need……Great need” in order to clarify its purpose to respondents; and

• The scales on the key items have been changed from seven-point to five-point scales, to reflect the preferences of respondents and to make the forms easier to use for visually impaired respondents using screen readers.

The respondents suggested a few changes in wording to improve understanding and suggested the addition of a few sub-items. Respondents also identified selected issues with the Web-based versions of the forms, including items that are difficult to use by those with visual impairments using screen readers. These changes have been integrated into the instruments submitted in this package.

B5. Persons Responsible for Design, Data Collection and Analysis

A contractor, Development Associates, Inc. of Arlington, Virginia is conducting the needs assessment. The Project Director for the evaluation contract is Dr. Malcolm Young (703-276-0677). Mr. Charles Sadler of the Rehabilitation Services Administration (202-245-7514) is managing the project.

2