William Yates, Trilby VanDeusen, Sierra Schuman, Keanna Curry, Josh Maki, Matt Ogborn, Allison Leaming Malecki, Joyce Ann Martin, Wes Edens, Mimmo Bonanni and Eric Prosser
This study aims to explore the use of artificial intelligence (AI) personas for training library staff in an online chat reference environment, enhancing trainees’ abilities to…
Abstract
Purpose
This study aims to explore the use of artificial intelligence (AI) personas for training library staff in an online chat reference environment, enhancing trainees’ abilities to respond effectively to real-life patron requests.
Design/methodology/approach
The project used ChatGPT-4 to construct custom generative pretrained transformers (GPTs) that acted as library patrons and presented the types of questions and reference requests that library staff would likely receive. These AI personas were improved iteratively and collaboratively. Novice chat reference users were surveyed on their experiences with the final AI personas.
Findings
The use of AI personas for training library staff showed promising results. AI personas could simulate interactions with actual patrons, and trainees’ confidence levels increased due to their use. However, limitations still exist, as AI personas were prone to providing overly detailed answers or required prompting to stay on task.
Originality/value
To the best of the authors’ knowledge, this study is among the first to apply AI personas in the context of improving training for online chat reference. The experience contrasts with other efforts to use generative AI to provide answers directly to patrons without human intervention or involvement.
Details
Keywords
Jessica K. Hardy, Katerina Marcoulides, Jill Grifenhagen, Ragan H. McLeod and Mary Louise Hemmeter
We developed and studied an approach to measuring the quality of coaching meetings. Coaching is a professional development approach that has been implemented in education settings…
Abstract
Purpose
We developed and studied an approach to measuring the quality of coaching meetings. Coaching is a professional development approach that has been implemented in education settings for several decades to support teachers and other practitioners in providing effective instruction. As coaching has become more prevalent, it has become clear that the field needs tools to measure coaching quality.
Design/methodology/approach
The coaching quality checklist (CQC) is a measure based on the empirical and theoretical literature on coaching. It has 26 items designed to measure three constructs: foundational, supportive and change-oriented coaching skills. In this study, we conducted a confirmatory factor analysis of the CQC.
Findings
We found the one-factor model fit the data well. The hypothesized higher-order three subfactor model fit the data better but not significantly so. Additional research is needed to further validate the CQC using a larger sample and examine different types of validity.
Originality/value
The CQC is a promising tool for measuring coaching quality, which can help ensure that teachers are provided with high-quality professional development to support their use of interventions.