Justification part B

Section B. Description of Statistical Methodology

B. 1. Respondent Universe

The PLS is a universe survey of public libraries that are identified by state library agencies (about 9,000) in the 50 states, the District of Columbia, and the outlying areas. State library agencies collect the data from their local public libraries, and state data coordinators employed by the agencies report the data to IMLS via the Web application by the Census Bureau. All 50 states, the District of Columbia participated in the FY2005 PLS. A total of 8,955 of the 9,214 public libraries in the survey universe responded to the survey, for a unit response rate of 97.2 percent.

B.2. Statistical Methodology

Not applicable, as this is a universe survey.

B.3. Methods for Maximizing Response Rate

The PLS has achieved a unit response rate by local public libraries of at least 96 percent since 1989. To maximize response rate and reduce respondent burden, the PLS was developed as an electronic data collection instrument. The web survey application includes a universe maintenance function, data editing/edit report tool, table generator, and user guide. These features were designed to provide maximum assistance to the respondent in completing the survey, to improve data quality, data timeliness, and to minimize edit follow-up for data problems.

To further reduce response burden, the survey contains pre-entered prior-year data for items that are not expected to change annually—about 35 percent of the survey items. The respondent is requested to review the pre-entered data and update it, if necessary.

For 2007, IMLS anticipates providing training for the new state data coordinators. In 2008, IMLS anticipates hosting a professional development conference to train all of the state data coordinators on the functions and use of the survey application. IMLS and the Census Bureau also would offer technical assistance to states that have difficulty in submitting their data due to staff shortages or other reasons. Staggered survey due dates were established to enable state data to be submitted closer to the end of the state fiscal cycle. This results in timelier data submissions and enhances the flow of data processing at the state and federal levels. PLS Steering Committee members serve as mentors to other state data coordinators. States with low item response rates are identified during the edit follow-up process and encouraged to improve their response rates.

B.4. Tests of Procedures and Methods

The Census Bureau developed the web application. The Census Bureau would be responsible for the web application’s testing and maintenance, with technical assistance and review to be performed by IMLS. The survey application would be tested by Census Bureau and IMLS staff members.

The PLS definitions are reviewed on an ongoing basis by the Steering Committee and revised when necessary to improve data quality. Survey items may be discontinued due to changing importance or poor data quality. Evaluations of the validity and reliability of selected data are made periodically, including recommendations for improving the data collection.

B.5. Reviewing Statisticians

Terri Craig, a mathematics statistician for the Census Bureau, reviews the data and is consulted on technical aspects of the data collection. In addition to Ms. Craig’s review, the survey statistician at IMLS will conduct a final data review prior to release of the data and report.

1