No abstract

In education, intelligent learning environments allow students to choose how to tackle open-ended tasks while monitoring performance and behavior, allowing for the creation of adaptive support to help students overcome challenges. Timely feedback is critical to aid students' progression toward learning and improved problem-solving. Feedback on text-based student responses can be delayed when teachers are overloaded with work. Automated evaluation can provide quick student feedback while easing the manual evaluation burden for teachers in areas with a high teacher-to-student ratio. Current methods of evaluating student essay responses to questions have included transformer-based natural language processing models with varying degrees of success. One main challenge in training these models is the scarcity of data for studentgenerated data. Larger volumes of training data are needed to create models that perform at a sufficient level of accuracy. Some studies have vast data, but large quantities are difficult to obtain when educational studies involve student-generated text. To overcome this data scarcity issue, text augmentation techniques have been employed to balance and expand the data set so that models can be trained with higher accuracy, leading to more reliable evaluation and categorization of student answers to aid teachers in the student's learning progression. This paper examines the text-generating AI model, GPT-3.5, to determine if prompt-based text-generation methods are viable for generating additional text to supplement small sets of student responses for machine learning model training. We augmented student responses across two domains using GPT-3.5 completions and used that data to train a multilingual BERT model. Our results show that text generation can improve model performance on small data sets over simple self-augmentation.

Simulation-based training (SBT) programs are commonly employed by organizations to train individuals and teams for effective workplace cognitive and psychomotor skills in a broad range of applications. Distributed cognition has become a popular cognitive framework for the design and evaluation of these SBT environments, with structured methodologies such as Distributed Cognition for Teamwork (DiCoT) used for analysis. However, the analysis and evaluations generated by such distributed cognition frameworks require extensive domain-knowledge and manual coding and interpretation, and the analysis is primarily qualitative. In this work, we propose and develop the application of multimodal learning analysis techniques to SBT scenarios. Using these analysis methods, we can use the rich multimodal data collected in SBT environments to generate more automated interpretations of trainee performance that supplement and extend traditional DiCoT analysis. To demonstrate the use of these methods, we present a case study of nurses training in a mixed-reality manikin-based (MRMB) training environment. We show how the combined analysis of the video, speech, and eye-tracking data collected as the nurses train in the MRMB environment supports and enhances traditional qualitative DiCoT analysis. By applying such quantitative data-driven analysis methods, we can better analyze trainee activities online in SBT and MRMB environments. With continued development, these analysis methods could be used to provide targeted feedback to learners, a detailed review of training performance to the instructors, and data-driven evidence for improving the environment to simulation designers.

No abstract