Computing scientists generally agree that abstract thinking is a crucial component for practicing computer science.

We report on a three-year longitudinal study to confirm the hypothesis that general abstraction ability has a positive impact on performance in computing science.

Abstraction ability is operationalized as stages of cognitive development for which validated tests exist. Performance in computing science is operationalized as grade in the final assessment of ten courses of a bachelor's degree programme in computing science. The validity of the operationalizations is discussed.

We have investigated the positive impact overall, for two groupings of courses (a content-based grouping and a grouping based on SOLO levels of the courses' intended learning outcome), and for each individual course.

Surprisingly, our study shows that there is hardly any correlation between stage of cognitive development (abstraction ability) and final grades in standard CS courses, neither for the various group-ings, nor for the individual courses. Possible explanations are discussed.