In this paper, we discuss the visual programming approach to teaching introductory programming courses and then compare this approach with that of procedural programming. The involved cognitive levels of students, as beginning students are introduced to different types of programming concepts, are correlated to the learning processes of programming. Our hypothesis is that if beginning students are introduced to programming concepts by means of a console-based procedural programming approach, they perform better in subsequent visual programming higher level courses. The performance of two groups of students, one group who began with the console-based procedural programming approach and then advanced to a higher level visual programming course and the other group which began with a lower level visual programming course before proceeding to the same higher level visual programming course, is measured, analysed, and the results are reported, with statistical analysis, and correlated to the hypothesis.