Computational Thinking (CT) skills are increasingly recognized as essential for junior high school students, especially in addressing the demands of the digital era. This study explores how CT skills--decomposition, pattern recognition, abstraction, and algorithmic thinking--manifest in learning statistics based on students' cognitive abilities. A qualitative research method was employed, involving 30 junior high school students, with six participants representing high, medium, and low initial abilities. This study uniquely maps students' CT performance in solving statistical problems, a domain that has been underexplored in relation to these skills. The results reveal significant differences based on cognitive ability: (a) students with high cognitive abilities demonstrate mastery of CT skills across all four indicators when solving statistical problems; (b) students with moderate abilities show partial competence, excelling in decomposition and abstraction but struggling with pattern recognition and algorithmic thinking; (c) students with low abilities achieve limited success, excelling in decomposition but facing challenges with the other CT skills. The novelty of this research lies in its focused examination of the intersection between CT skills and statistical problem-solving in junior high students, offering valuable insights for curriculum development. The findings suggest that integrating CT skills into statistics education enhances problem-solving capabilities across varying cognitive levels, promoting more inclusive and effective learning in the digital era.