Introduction.   In 2006, while in graduate school at the Kent State University School of Library and Information Science, I completed an internship under Dr. Neal Kaske, who, at that time, served as Director of Statistics and Surveys of the U.S. National Commission on Libraries and Information Science (NCLIS). (Dr. Kaske is currently Director of the U.S. National Oceanographic and Atmospherice Administration Libraries.) The project included a statistical analysis of National Center for Educational Statistics (NCES) U.S. public libraries survey data for 1999 - 2004 (54,824 records) and an examination of the Hennen American Public Library Ratings (HAPLR).

Findings from this research appear in "Unsettling Scores: An Evaluation of the Hennen Annual Public Library Ratings" Public Library Quarterly 26 (3-4), Winter 2007, pp. 49-100 (12MB PDF viewable here). This article references the larger set of charts prepared for the project, which appear in my graduate thesis and are also posted here. Links to the charts appear below.

Statistical Contents of the Charts.  The data studied are the traditional library measures that have been utilized since the early 1900's, such as circulation, visits, expenditures, and so on. These traditional measures were the focus of the HAPLR methodology; so the study did not look at more modern measures, such as electronic materials use or library programming.

For purposes of this study the data were combined from six reporting years, 1999 through 2004. Rather than describing any given year, or trends across years, the research objective was to identify general patterns in library measures that might shed light on their potential use in rating systems such as HAPLR. (Combining years 'smooth outs' any patterns in the data due to to single-year changes in data collection or reporting or to causes.) This was important for analyzing basic aspects of the measures, in particular their distributions (ranges, quartiles, medians. etc.)

Given the research objective, I chose to render the data in a style known as "box-and-whisker plots" or "boxplots". This graphical plotting method is quite useful for general data exploration. Prior to viewing the charts, readers should refer first to Understanding the Contents of the Study Charts. Note also the link there entitled Guide to Interpreting Boxplots.

 

Click on any item link to view charts for that statistical item. Or use the <Browse Charts> link at the top of this page to browse through the charts in order.