<?xml version="1.0" encoding="UTF-8"?>

<metadata
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xmlns:dc="http://purl.org/dc/elements/1.1/"
xmlns:dcterms="http://purl.org/dc/terms/">
<dc:title>Average Distance to Public Libraries by State</dc:title>
<dc:creator>Frank Donnelly</dc:creator>
<dc:subject xsi:type="dcterms:LCSH">Public libraries--United States</dc:subject>
<dc:subject xsi:type="dcterms:LCSH">Public libraries--Statistics</dc:subject>
<dc:description>These data tables were created as part of a project that is summarized in the following paper:

Francis P. Donnelly, Regional variations in average distance to public libraries in the United States, Library and Information Science Research, Volume 37, Issue 4, October 2015, Pages 280-289, http://dx.doi.org/10.1016/j.lisr.2015.11.008.

If you use this dataset in conjunction with the article, cite the article. If you use the dataset by itself, you can cite it as follows:

Donnelly, F.P. (2016, February 9). Average Distance to Public Libraries by State. Retrieved from: http://gothos.info/resources/.

The data is released under a Creative Commons Attribution-Non Commercial-No Derivatives 4.0 International License. Every effort was made to insure that the data, which was compiled from public sources, was processed accurately. The author disclaims any liability for errors, inaccuracies, or omissions that may be contained therein or for any damages that may arise from the foregoing. Users should independently verify the accuracy of the data for their purposes.

Data in these tables was created using population centroids for census block groups from the 2010 Census and the location of public libraries from the IMLS 2010 fiscal year public data files. The coordinates of the population centroids of the 216,864 census block groups were plotted, along with their population from the 2010 Census, in a geographic information system (GIS). Library addresses in the IMLS dataset that lacked longitude and latitude coordinates that were accurate to an address or street segment were re-geocoded to get an improved match. As a result, 95% of the public libraries had X and Y coordinates that were accurate to an address or street segment while the remaining 5% represented ZIP Code centroids. Bookmobiles, books by mail centers, and libraries in US territories were excluded. The 16,720 public libraries were plotted in GIS, and the coordinate reference system for both the population centroids and libraries was converted from NAD 83 to North American Lambert Conformal Conic. GIS was used to measure the straight-line, Euclidean distance from each block group population centroid to the nearest public library.

Python scripts were written to process the resulting data table. Distance measurements were converted from meters to miles, and in the first table (states2010_publibs_totpop) the population of each block group was summed by state and by distance in one-mile increments - number of people living less than one mile from the closest library, from one to two miles, etc. The last category was coded to include all people living more than six miles from a library. A second table (states2010_publibs_pctpop) was generated to calculate the percent of each state's total population that lived within the one-mile increments. 

A third table (states2010_publibs_avgdist) was created to include summary statistics: the mean and median distances to each public library for each state, along with the range, standard deviation, and a count of the number of block groups. A population weighted average was also generated to account for the fact that the population of block groups varies (between 600 and 3,000 people): the distance for each block group to the closest library was multiplied by the block group's population, and the sum of the result was divided by the total population of each state to obtain a weighted-arithmetic mean. Block group centroids that had zero population were excluded from all summary statistics.

The tables are provided in two formats. The formats ending with .txt are text tables that are formatted for identifying individual figures and for printing. The formats ending with .csv are intended for use in spreadsheet or statistical programs. </dc:description>
<dcterms:created xsi:type="dcterms:W3CDTF">2013-06-24</dcterms:created>
<dcterms:available xsi:type="dcterms:W3CDTF">2016-02-09</dcterms:available>
<dc:type xsi:type="dcterms:DCMIType">Dataset</dc:type>
<dc:format xsi:type="dcterms:IMT">text/csv</dc:format>
<dc:format xsi:type="dcterms:IMT">text/plain</dc:format>
<dc:identifier>http://gothos.info/resources/</dc:identifier>
<dc:source>Centers of Population (2010) Census block groups</dc:source>
<dc:source xsi:type="dcterms:URI">https://www.census.gov/geo/reference/centersofpop.html</dc:source>
<dc:source>Institute of Museum and Library Services (2010). Public libraries (public use) data files</dc:source>
<dc:source>http://www.imls.gov/research/pls_data_files.aspx</dc:source>
<dc:language xsi:type="dcterms:ISO639-2">en</dc:language>
<dcterms:isReferencedBy>Regional variations in average distance to public libraries in the United States</dcterms:isReferencedBy>
<dc:relation>http://dx.doi.org/10.1016/j.lisr.2015.11.008</dc:relation>
<dcterms:temporal>2010</dcterms:temporal>
<dcterms:spatial xsi:type="dcterms:Box">name=United States;northlimit=71.6;eastlimit=-66.1;southlimit=18.6;westlimit=-178.6;units=degrees;projection=NAD83;</dcterms:spatial>
<dcterms:license>Creative Commons  Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)</dcterms:license>
<dc:rights>http://creativecommons.org/licenses/by-nc-nd/4.0/</dc:rights>
<dcterms:rightsholder>Francis P. Donnelly</dcterms:rightsholder>

</metadata>
