Purpose & Goals

Considering our university community’s evolving research needs and the changing demands of our branch library’s research and instruction team, the author set out to better understand the research interactions that take place at a shared service point where other library staff and undergraduate students also provide circulation and public service support. The author wanted to know answers to the following questions: how many and what types of research questions do we receive through our library’s shared service point, when do they typically come in, and what type of staff member fields them? The author incorporated responses from generative AI tools to enhance the data analysis and visualization portion of the project. This poster will highlight some of the potential advantages and challenges associated with using such tools in their current state, especially as it relates to the data novice.

Design & Methodology

Materials include data recorded in LibInsight over the course of 1 academic year, LibInsight visualization tools, ChatGPT-3.5, Google Sheets Methods: The author analyzed over the 2023-2024 academic year. This data came from library staff, graduate students, and undergraduate student workers who input research interaction information into a LibInsight dataset. Relying solely on Libinsight visualization tools proved ineffective at answering the specific questions motivating this assessment work. Therefore, the author pursued additional and more in-depth methods of analysis outside of the available system. The author exported the data for analysis and visualization, using generative AI chat bots to enhance the kind and variety of functions employed in Google Sheets. The author extracted quantitative and qualitative information from the dataset and used that to draw initial conclusions.

Findings

Preliminary results: We addressed 173 research interactions through our shared service point and other unscheduled means. 50% of research questions we fielded through unscheduled interactions dealt with basic library search tool navigation. Resource access issues, referrals to other services, and citation support each constituted less than 15% of all interactions. Undergraduate student staff fielded 40% of unscheduled research desk interactions, 50% by full-time staff, and 10% by graduate student staff. 65% of interactions took place when research and instruction team members were staffing the desk, while 25% occurred after research and instruction team members availability ended at the shared service point.

Action & Impact

Although I oversee research services for our branch library, I plan to bring my preliminary findings and recommendations into dialogue with stakeholders, namely the rest of the public services team (research and instruction; circulation, our branch library’s director). These conversations will influence adjustments to our data collection workflow, potential changes and developments for our research service staffing model. I am actively exploring opportunities to reduce graduate assistant time on the desk due to the lack of interactions fielded during their availability and the basic level of the questions coming in (simple search navigation and troubleshooting e-resources).

Practical Implications & Value

Other members of the community may also be grappling with how and whether to staff a research desk for walk-up interactions. The considerations drawn from the data analysis may help their own thoughts and action plans towards investigating their current offerings. Similarly, community members may benefit from learning methods that incorporate formulas and visualization strategies gleaned from generative AI chatbot prompts.