An R Package with a Shiny Dashboard for Visualizing and Comparing Library Statistics Data from Association of Research Libraries
libraryStatistics is an R package with a Shiny dashboard that permits
to visualize and compare data from the annual survey of Association of
Research Libraries (ARL; URL: www.arlstatistics.org/data/main). ARL data
describes the collections, staffing, expenditures, and service
activities of the ARL member libraries in the United States and Canada.
This R package is designed for the analysis and visualization of library
statistics published from the annual survey. Both the R package and
Shiny application enables the generation of statistical ratios for
comparative analyses. Through the interactive interface of the Shiny
application, users can dynamically visualize ratios by selecting ARL
member libraries, years, and also by creating custom ratios, which
enhances the usability of the published ARL data. Library statistics
data published from the annual ARL survey can be downloaded from ARL
Data Portal for any number of years and across any number of ARL member
libraries. However, at one time, both the R package and Shiny
application would only enable to perform analysis on 5 ARL member
libraries and 5 distinct years. The libraryStatistics package was
developed using R version 4.3.2 (2023-10-31),
Platform: x86_64-apple-darwin20 (64-bit) and
Running under: macOS Ventura 13.2.
To install the latest version of the package:
require("devtools")
devtools::install_github("anjalisilva/libraryStatistics", build_vignettes = TRUE)
library("libraryStatistics")To run the Shiny application:
libraryStatistics::shinyLibStats()To list all functions available in the package:
ls("package:libraryStatistics")
data(package = "libraryStatistics")
browseVignettes("libraryStatistics>")libraryStatistics contains 8 functions.
- shinyLibStats opens the Shiny application/dashboard which permit to perform visual comparisons of up to 5 ARL member libraries and up to 5 years, at one time, from user uploaded ARL annual survey data.
- visTotalLibraryExp permits to visualize total library expenditures in United States Dollars (USD) as ratios in comparison to various statistics reported in the annual survey of ARL as bar plots. This ratio shows question 6 on ARL survey as the numerator.
- visTotalLibMaterialsExp permits to visualize total library materials expenditures in United States Dollars (USD) as ratios in comparison to various statistics reported in the annual survey of ARL as bar plots. This ratio shows question 7 on ARL survey as the numerator.
- visProfStaffSalaries permits to visualize salaries of professional library staff in United States Dollars (USD), as ratios in comparison to various statistics reported in the annual survey of ARL as bar plots. This ratio shows question 8a on ARL survey as the numerator.
- visProfStaffCounts permits to visualize library professional staff counts, full-time equivalent (FTE), as ratios in comparison to various statistics reported in the annual survey of ARL as bar plots. This ratio shows question 13a on ARL survey as the numerator.
- visSupStaffCounts permits to visualize library support staff counts, full-time equivalent (FTE), as ratios in comparison to various statistics reported in the annual survey of ARL as bar plots. This ratio shows question 13b on ARL survey as the numerator.
- customRatioBuilder permits to build and visualize a custom ratio based on user selected numerator and denominator from various statistics reported in the annual survey of ARL.
- indexTableGenerator permits to build a table containing ARL Investment Index over years as reported in the annual survey.
An overview of the package is illustrated below:
The R package and Shiny dashboard permit to visualize, track trends, and compare data downloaded directly from the ARL data portal (www.arlstatistics.org/data/main), with no data cleaning involved. At one time, up to 5 ARL member libraries and 5 distinct years can be compared using this tool. The tool is designed for the analysis of library statistics published from the annual survey conducted by the ARL. The R package contain functions that permit the user to read in data downloaded from the ARL Data Portal and perform visual comparisons using ratios. The input data file should be in comma-separated value (.csv) format as directly downloaded from the ARL Data Portal. Users must ensure that all variables are selected when downloading data, with columns set to ‘Variables’ and the data sorted by ‘Institution Name’ (which are the default options on ARL Data Portal). As such the first column must be labeled ‘Year’, followed by other indicators in any order, such as ‘Institution Name’, ‘Institution type’, etc., as directly downloaded from the ARL Data Portal. Data may be downloaded for any number of years and across any number member libraries/institutions (or all) available.
For the Shiny application, the user is able to upload a dataset in .csv format that is directly downloaded from ARL Data Portal (as explained above). Upon uploading data, the choices for libraries and years will be visible. Through the interactive interface of the Shiny application, users can dynamically visualize ratios by selecting up to 5 ARL member libraries and up to 5 years.
Functions of the R package and dedicated tabs of the Shiny dashboard are available for generating visualizations for ARL Investment Index, ratios for ‘Total Library Expenditures’, ‘Total Library Materials Expenditures’, ‘Professional Staff Salaries’, ‘Professional Staff Counts’, and ‘Support Staff Counts’. Some examples of ratios that can be visualized are total library expenditures (in USD) per faculty, per student, and per doctoral degree awarded, etc. Users also have the option to generate and visualize custom ratios using all available ARL indicators, which enhances the usability of the published ARL data.
Tool is useful in helping ARL libraries understand their own performance, benchmark performance against other ARL member libraries, and to analyze overall research library trends in North America. The tool does not hold any data, which respect data privacy regulations. Further, the tool is useful in detecting data anomalies and improving errata practices. Overall, the tool enhances the utilization of ARL collected data in making evidence-based decisions within UTL and other libraries, to gain insights into the multifaceted ways in which library resources contribute to its community, to support research and scholarship endeavors.
Note: corrections made by a library in the ARL survey, after data submission deadline, will be reflected in the footnotes of future years, but the dataset will not be updated to reflect the corrections. Be mindful of this when visualizing data downloaded from ARL Data Portal.
The Shiny application employing libraryStatistics could be run and results could be visualized using following command:
libraryStatistics::shinyLibStats()
In simple, the shinyLibStats is a web application available with
libraryStatistics.
For tutorials and plot interpretation, refer to the vignette:
browseVignettes("libraryStatistics")citation("libraryStatistics")Silva, A. and K. Maidenberg (2024). libraryStatistics: An R Package with a Shiny Dashboard for Visualizing and Comparing Library Statistics Data from Association of Research Libraries. Unpublished.
A BibTeX entry for LaTeX users is
@misc{,
title = {libraryStatistics: An R Package with a Shiny Dashboard for Visualizing and Comparing Library Statistics Data from Association of Research Libraries},
author = {A. Silva and K. Maidenberg},
year = {2024},
url = {https://github.com/anjalisilva/libraryStatistics},
}-
Association of Research Libraries. (2023). ARL Statistics 2023 Instructions.
-
Wickham, H. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York, 2016.
-
Bache, S., Wickham, H. (2022). magrittr: A Forward-Pipe Operator for R. R package version 2.0.3.
- Anjali Silva ([email protected]).
libraryStatistics welcomes issues, enhancement requests, and other
contributions. To submit an issue, use the GitHub
issues.
- Access to ARL annual survey data provided by University of Toronto Libraries (UTL), Ontario, Canada. We wish to thank Larry Alford, Laura Anderson, Glen Morales and Caitlin Tillman from UTL and Dr. Kevin Borden and Holly Gross from ARL for their feedback.