Exploring the impact of the uncanny valley effect on teenagers’ attitudes toward virtual characters and EFL reading comprehension skills

This study investigates the impact of the uncanny valley (UV) phenomenon, originally introduced by Masahiro Mori in 1970, on reading comprehension among Turkish secondary school students using interactive e-books. The UV concept describes the discomfort people experience when encountering entities that appear almost human but have unnatural features. A total of 86 students, aged 10–15, participated in this quantitative research, where they engaged with characters exhibiting varying degrees of human likeness in English-voiced short stories. A Reading Comprehension Test (RCT) assesses reading comprehension of English as a Foreign Language (EFL). The RCT included 15 stories sourced from K5 Learning. This educational platform provides resources to enhance reading comprehension for kindergarten to grade 5 students, focusing on skills like identifying context clues, understanding cause and effect, and determining main ideas. The Attitude Toward Robots and Robot-Related Experiences Scale was also used to measure students’ attitudes toward computer-generated morphs. At the same time, the uncanny valley Scale assessed participants’ reactions to five computer-generated models to determine the presence of the UV effect. The study’s descriptive and inferential statistical analysis revealed that hyper-realistic characters elicited stronger feelings of eeriness, reducing engagement and negatively impacting comprehension. Heat maps further identified the eyes and lips as key areas triggering the UV effect, underscoring the importance of careful character design in educational tools. The study found that the UV effect significantly influenced participants’ EFL reading comprehension scores, particularly when characters were designed to evoke it. Characters in the ‘uncanny’ range were found to diminish engagement and comprehension, suggesting that a balance between realism and stylization in avatars is crucial for enhancing engagement and minimizing negative emotional responses. These results implicate the importance of visually balanced designs in educational technology to improve student engagement and learning outcomes.