This study investigates the emotional and cognitive responses of Arabic-speaking audiences to informal science communication on YouTube. Focusing on three prominent Arabic YouTube channels that provide science content, @Da7ee7, @NidhalG, and @Espitalia, the study analyzes their communication styles to explore how stylistic differences shape audience engagement and perception. The study classifies the channels into three distinct communicative styles; humorous, academic, and conversational to examine variation in audience responses. Employing computational content analysis, the study applies LDA topic modeling and sentiment analysis to examine emotional and cognitive engagement in user comments. Results show that humorous content is associated with stronger positive emotions and cognitive activation, whereas academic and conversational styles evoke more mixed reactions. Topics like space science and artificial intelligence elicit complex emotions such as confusion and admiration. Addressing a key gap in Arabic-language science communication, the study reveals how style and topic shape public engagement in culturally specific digital spaces.