EMOTIONS AND LEARNING IN EARLY CHILDHOOD: EXPLORING PRESCHOOLERS’ VISUAL SCENE COMPREHENSION USING AI-GENERATED IMAGES
University Paris-Est Créteil (FRANCE)
About this paper:
Conference name: 18th International Conference on Education and New Learning Technologies
Dates: 29 June-1 July, 2026
Location: Palma, Spain
Abstract:
From an early age, children encounter emotionally meaningful information in their daily environments, both at home and in educational settings. In preschool, emotional development is actively supported as children are guided to recognize, experience, and express their emotions. Research in cognitive psychology has long emphasized the close relationship between emotion and cognition, showing that emotional processes play a key role in shaping judgment, decision-making, and learning. Despite this growing body of work, limited research has explored how emotional content affects both surface-level comprehension (memorization) and inferential comprehension in young pre-literate children.
The present study investigates the influence of emotional valence on the comprehension of visual scenes in preschool children. Thirty-four participants (mean age = 5.15 years) took part in an experimental task. Each child viewed three AI-generated visual scenes depicting a fictional character within an environment. The character displayed one of three emotional valences: neutral, positive, or negative. The order of presentation was randomized. After viewing each image for five seconds, participants evaluated its emotional valence using a five-point Self-Assessment Manikin (SAM) Likert scale and answered questions assessing surface comprehension and causal inference.
Results indicate that participants successfully perceived the emotional valence of the presented images. Comprehension scores suggest that positively valenced images led to higher inferential-level comprehension scores than negatively valenced images. Cognitive and pedagogical implications for early learning environments are discussed, as well as the potential of AI-based image generation as a tool for producing emotionally controlled visual materials for research and educational purposes.Keywords:
Emotion, learning, AI-based image, preschoolers.