On Wednesday, Dec. 4, Library Journal hosted the first of a three-part series on using data to drive decisions in a library. The webcast is available for viewing, and here’s a summary.
Sarah Tudesco, Assessment Librarian at Yale University, organized her talk around the question of the percentage of materials used by patrons.
The basic outline of a data-driven decision: Question, Plan, Collect, Analyze, Recommend
1. Begin by developing a question: Who is using the library?
Then brainstorm smaller questions. Examples:
Usage trending up & down? Do UG students use it? What services do they use? Spaces used heavily? by whom? What spaces underutilized?
What is the library not providing?
Need to have a solid grasp of the questions you’re trying to answer. Questions anchor you and help you discover the core elements.
2 & 3. Plan/Collect.
Plan: Refine the question (data source(s))
Who is coming to the library (Space)
Who is checking out books (ILS)
Who is using electronic resources (Web analytics)
Collect: Look for data sources
Space: turnstile data, head-counts, room reservation systems
ILS: circulation reports
Web: web analytics, link resolver data, vendor usage reports
Systems data: collections, financials, circulation, electronic resources, resource sharing (ILL, consortial agreements), gate counts, web analytics, special collections (e.g. journal usage in Access)
Workflow sources: reference, bibliographic instruction, cataloging, preservation (tipping in pages)
Patron input: surveys, focus groups, social media analysis
3. Analyze. Take the time to look at data from many different angles, with different breakdowns.
Circulation by patron group – raw numbers
Circulation by patron group – as percentage of the total
Circulation by specifically undergraduates: Seniors vs. lower years
Monthly circulation vs. fiscal year
Think about impact
Outcomes of a change in service
Sarah then talked about skills librarians need to gather and analyze data:
Data Capture; Programming & database skills; Business domain knowledge; Data Analysis; Spreadsheets; Reporting tools; Statistics; Presentation; Visualization;
She mentioned tools librarians can use to gather and analyze data:
Spreadsheets (Excel, many more functions than Google)
Stats tools: R, SATA, SPSS
Survey Tools: Qualtrics, Webmonkey, Google Forms
Visualization: Tableau, Many Eyes
Sarah mentioned something that was new to me: “Hype Cycles” from the market analysis company Gartner Group. She showed a “Hype Cycle” for Emerging Technologies, including Big Data, as of 2012. (Link goes to 2013 Cycle)
Gartner organizes a Cycle into different periods: technology trigger, peak of inflated expectations, trough of disillusionment, slope of enlightenment, plateau of productivity (mainstream adoption). Plateau will be reached in (less than 2 years, 2-5 years, 5-10 years, 10+ years, obsolete before plateau
According to Gartner, Big Data is at the “peak of inflated expectations” stage. Gartner predicts this concept still has 5-10 years before it reaches mainstream adoption.