Teaching Programming to Science Majors by Modelling

Programming is playing an increasingly important role in various fields, including science. However, traditional programming instruction tends to use small-scale and general examples to explain syntax and semantic meaning of the code, which cannot foster students' programming ability of solving real-world problems. This research was intended to develop a modelling-based instruction for scientific programming to guide students to solve programming problems based on the modelling process (phenomenon description, data modelling, algorithmic modelling, coding, and verification and debugging). A learning platform based on the proposed modelling process was also developed to assist science-major students to learn how to solve real-world scientific problems by programming. An empirical study was conducted on thirty-two science-major college students to prove the effectiveness of the modelling-based scientific programming. The experiment results show that students who engaged more in modelling had higher programming performance. The modelling-based instruction actually helps students to write programs for solving scientific problem by using both of data and algorithmic models.