Many frontline library workers are asked to count certain interactions with patrons, such as the number of reference questions they answer. At Library Research Service (LRS), we call these reference interactions between patrons and staff “reference transactions.” Being tasked with tracking this data might also naturally lead staff to wonder, “What exactly qualifies as a reference transaction?” and “Where does this data end up, and why is it important?” In this post, we’ll share some new guidelines from LRS on counting reference transactions, see what we can learn from the reference data we have collected, and share some advice for a more accurate reference count.
What Counts?
Let’s first address the not-so-simple question of what should and should not be included in a library’s count of reference transactions. LRS defines reference transactions as “information consultations in which library staff recommend, interpret, evaluate, and/or use information resources to help others to meet particular information needs.” This broad definition leaves room for different interpretations. People may not have the same idea of what counts as a consultation or what can be labeled an information resource. Below is a summary of the new clarifying guidelines that LRS released this year to limit any confusion about what to report. For the exact instructions, please refer to our Colorado Public Library Annual Report Question Descriptions document, and if you have additional questions, please reach out!
What a reference transaction IS:
- Reference transactions are scheduled appointments OR spontaneous interactions where individual instruction or assistance is given using an information source.
- Information sources that could be used in reference transactions include:
- printed and nonprinted material.
- machine-readable databases (including computer-assisted instruction).
- the library’s own catalogs and other holdings records.
- other libraries and institutions through communication or referral.
- persons both inside and outside the library.
- Reference transactions can include assistance using information sources such as computers or specific websites.
- Readers’ Advisory questions are reference transactions.
- If a staff member answers a question using information previously found in an information source without consulting that source again, it is still a reference transaction.
- The amount of time put into answering a reference question in no way influences whether it should be counted as a reference transaction.
What a reference transaction is NOT:
- Directional questions, such as where certain staff, library users, material types or bathrooms are located, are NOT reference transactions.
- If a directional question is asked alongside a reference question. It should still be counted as one reference transaction.
- Assisting with equipment use that does not include an information source, such as helping someone make a photocopy, is not a reference transaction.
Why Count?
The short, oversimplified answer to why library staff are asked to count reference interactions is that it is a mandatory piece of data for us to gather in the Public Library Annual Report (PLAR). In other words, the number of reference transactions across Colorado is part of the data that we (along with every other state) are required to report to the federal government. However, taking a closer look at the local level, tracking reference transactions can give libraries crucial information about where to allocate resources and how best to serve their communities.
Although LRS is only required to collect and report on the number of reference transactions overall, some libraries choose to dive deeper and categorize the types of reference questions they answer. Recording how many questions are asked about social services (such as housing resources or employment assistance) can help demonstrate the value libraries bring to their communities. Sharing this data on reference transactions to demonstrate the breadth of information libraries provide and the number of times they assist patrons can also help libraries acquire the funds needed to continue providing these resources.
Understanding the information needs of the communities a library serves can help direct everything from collection development to programming. Tracking the number of reference transactions year-over-year can also help a library understand whether patrons are regularly engaging with the library as a trusted information source. As the landscape of information access has and will continue to shift drastically, libraries will likely see changes in their reference data. Tracking these changes as the availability of assistive technology, such as Artificial Intelligence (AI) tools, explodes can increase our understanding of how people are adapting and identifying credible sources.
Conducting a Careful Reference Count
In the PLAR, libraries are asked to report how they count reference transactions. Reference transaction data can come from either an annual count (CT) or an annual estimate based on a typical week (ES). In 2024, 43% of reporting Colorado public libraries used the annual count method, meaning that they kept track of how many reference transactions took place throughout the entire year. Over half of libraries (57%) reported that they make an estimate based on counting reference transactions only during certain time periods. If a library counts reference transactions throughout an entire year it is likely an integrated part of staff workflow. But this method is not feasible for every library. For libraries that choose a week or multiple weeks to count reference transactions, this task can feel like a significant burden on top of their usual workload. However, the count collected during this time will be multiplied to approximate the annual total reported on the PLAR, so counting accurately during this period is the only way to ensure a close estimate. We suggest that libraries conduct this count when:
- The library is historically neither busy nor slow (avoid holidays).
- The library is open typical hours.
- There are not any unusual events taking place in the community.
- Key staff are present (not on leave).
A small library can more easily identify when key staff will be present to conduct the count and communicate to staff members the importance of this data. Larger systems may have more difficulties coordinating a count across branches. One of the best ways to improve the accuracy of a reference count is to encourage buy-in from the staff recording their reference transactions with patrons. Frontline library staff are incredibly busy and may face an untold number of challenges as they interact with patrons throughout the day. If this to-do item is added to their task list without an explanation of why counting reference transactions is important, where this data goes, or how it is used, it likely won’t be considered a top priority. We hope this post can help by sharing the impact this data can have and giving clear instructions on what to count as a reference transaction to limit any frustration around this data collection, leading to more accurate counts across Colorado public libraries.
Compiling and Sharing Reference Data
With over half of libraries across the state estimating the number of reference transactions they report, it’s safe to say that the PLAR data on reference transactions are not exact. It falls into the category of what we at LRS like to call “squishy” data. Rather than focusing on specific counts, one practical use for this estimated data is to examine trends over time. However, now that LRS has published clarifying guidelines for counting reference transactions, it is likely that the data will be counted slightly differently moving forward. In 2024, the data was counted as it previously had been since these new guidelines had not yet been communicated. Let’s take a look at how reference transactions fluctuated across the state over the past six years.
Unsurprisingly, total reference transactions across the state fell drastically during COVID-19. From pre-COVID-19 (2019) to 2020 transactions fell 62%. This decrease is similar, though not quite as drastic, to that of library visits, which fell 68%. In 2021, reference transactions jumped back up to 63% of what they had been in 2019, while visits took a longer time to start recovering, staying at less than half of what they were in 2019. This difference in recovery could be attributed to the fact that reference transactions can also take place over phone, chat or email. In 2024, reference questions bounced back to 78% and visits had recovered to 73% of the 2019 totals. Although reference transactions and library visits across the state were reported at around three-fourths of their pre-COVID-19 totals in 2024, LRS is happy to share that circulation in 2024 was back up to 96% of what it was in 2019!
The reality of working with real-world data is that it is often much messier than we might like. By setting these new guidelines for reference transactions we hope to increase consistency in the data recorded across Colorado. Additionally, when compiling this data at LRS, we have quality controls in place to help catch errors by flagging data that doesn’t fall within certain parameters. This is part of the data cleaning process. Messy data requires more cleaning and should be used cautiously. If the data was estimated, it’s important to be transparent about this when sharing it with stakeholders. Even though it might be “squishy,” data on reference transactions can be used carefully and combined with other observations to support decision-making. Comparing totals from year to year can help shine a light on libraries’ evolving role in information access.
LRS’s Colorado Public Library Data Users Group (DUG) mailing list provides instructions on data analysis and visualization, LRS news, and PLAR updates. To receive posts via email, please complete this form.