This work-in-progress paper is part of a project that aims to analyze how the implementation of different instructional designs through complete and incomplete examples influence the cognitive load that students perceive when learning statistics and computer programming. For this first stage, we conducted a pilot study aimed at conducting a differentiated measurement of cognitive loads through secondary tasks and a subjective scale. The pilot study consisted of three moments. The first one focused on measuring their working memory capacity and their prior knowledge of statistics. Then, students worked on a learning activity focused on using Chi-square and Pearson correlation to conduct inferential analyses. The study included two different instructional designs in the form of computational notebooks. Design 1 consisted of correct and incomplete examples, while Design 2 included correct, incomplete, and incorrect examples. Simultaneously, students had to attend to a secondary task, which consisted of responding to an auditory stimulus, after which they had to press a key on their keyboard. Finally, students completed a naïve rating scale of cognitive loads involved in the task. The results highlight the important role of prior knowledge in different instructional designs and how it connects to students' perceived cognitive effort. Likewise, the differences in the instructional designs influenced students' cognitive loads, suggesting a differentiated measure of extraneous loads. Using incorrect examples in learning statistics and programming may affect students' perceived difficulty without resulting in a better learning outcome.