From logs to insights: Exploring domains of creativity in data analytics tasks

Abstract

Although creativity research has traditionally focused on observable outcomes, less attention has been given to the real-time behavioral and cognitive processes that underlie creative thinking. In this exploratory study, we investigate how creativity manifests in data analytics (DA) tasks - an increasingly important domain in both education and professional practice. By analyzing correlations between participants' self-reported creativity, as measured by the Kaufman Domains of Creativity Scale (K-DOCS) questionnaire, and their recorded behaviors as reflected in system log file data of DA platform, we found that different domains of creativity are linked to distinct behavioral patterns in DA tasks. The results show that among the five domains of creativity, the Scholarly domain was the most prominent. Participants with higher scholarly creativity scores achieved better outcomes in the DA task while using fewer actions, suggesting more efficient DA proficiency. These participants also used platform features such as “help” and "about" less frequently than others. The Artistic domain emerged as the second most prominent. Participants with higher levels of artistic creativity also completed the task successfully using fewer actions, implying that sensitivity to visual and structural cues may facilitate efficient DA proficiency. These findings contribute to the domain-specificity perspective on creativity by showing how distinct creativity profiles relate to different behavioral patterns in DA tasks. They also lay a foundation for developing less biased, more cost-effective, log-based assessment tools that can help educators and researchers identify and support creative thinking in DA tasks.