A pilot feasibility study of chatbot-guided self-reflection in golf putting

Self-reflection is a key psychological process for enhancing self-efficacy and sport performance, yet traditional reflection tools impose high cognitive demands and occur post-event when memory has faded. This pilot examined whether a chatbot delivering real-time, guided self-reflection could improve task self-efficacy and putting performance in golf. Fourteen golfers (12 male, 2 female, Mage = 37 years) completed five rounds of putting in a controlled indoor environment, engaging with the chatbot between rounds. Self-efficacy and putting accuracy were assessed pre- and post-intervention, and usability was evaluated using the System Usability Scale and qualitative feedback. Significant improvements from pre- to post-trial were found in self-efficacy magnitude (p = .020), self-efficacy strength (p = .007), and golf putting performance (p < .001). Mean usability was rated as 88.8 (SD = 7.39) out of 100, above the benchmark for “good” usability (M = 71). Participants’ highlighted the chatbot’s ease of use, structured prompts, and value for in-task reflection, while also suggesting refinements such as incorporating voice-to-text features and more golf-specific content. Overall, this pilotindicates that chatbot-guided reflection may offer a promising complement to traditional psychological skills training in sport, with further research needed to determine whether chatbot-delivered prompting contributes to changes in reflection, self-efficacy or performance beyond the effects of repeated practice.