Using AI to Create Personalized, Structured Training Plans for the World Marathon Majors
Training for a marathon requires discipline, structure, and support. Many people are able to achieve their goals by joining training groups, or getting advice from an online or an in-person coach. However, some runners lack access to training groups or coaches due to location, financial status (Smith, 2025), social predicament, or work schedule (HRP, 2024). Using Generative AI (GenAI) to create a structured, personalized training plan is an alternative to a human coach or training group. However, the quality of training plans created by GenAI can vary dramatically depending on the input (Düking, 2024). If runners do not know how to create accurate context-based prompts, the training plan will not be very useful. Therefore, the goal of this project was to design and evaluate instruction that assists runners in using GenAI to create structured, personalized marathon training plans that will guide them to their marathon goals.
The instructional design employed a mix of direct, indirect, and experiential learning strategies. A usability study with three participants was conducted, yielding largely positive feedback with 68% positive comments overall. Navigation was reported as intuitive with 82% positive comments, and content was comprehensive with 75% positive feedback. The primary usability concern was the Worksheet Functionality with 59% negative comments, which resulted in converting the original Google Form into a clearer, editable Google Doc. Learning effectiveness was measured with pre- and post-assessments by participants (n=15). The average score on the post-assessment (89%) showed a significant improvement of 21% over the pre-assessment (68%), demonstrating high instructional effectiveness. Furthermore, post-survey attitudinal results showed that 100% of participants would recommend the instructional website, and on a scale of 1 to 5, with 1 being not helpful and 5 being very helpful, participants reported an average score of 4.3 regarding GenAI’s usefulness for creating personalized training plans. These results suggest the instruction successfully met its goal of teaching middle-aged runners to leverage Generative AI for their marathon training.
Rosalie Paradise, LTEC Student, University of Hawaiʻi, US
Ruadhán Buddenhagen, LTEC Student, University of Hawaiʻi, US
Jimena Andrea Riano Tellez, LTEC Student, University of Hawaiʻi, US
Justice Kanaulu, LTEC Student, University of Hawaiʻi, US
Mariam Jafaari, Asian University for Women (AUW), BD
Maryam Mirzaee, Asian University for Women (AUW), AF
Corrin Barros, University of Hawaiʻi, US
Marufa Bhuiyan,