DOI
https://doi.org/10.1145/2676660

Summary
A qualitative study of five undergraduate biology majors in a data-driven computer science course is presented in this article. The authors introduce the students’ profiles and describe the data collection method. The data collection was performed in the form of MATLAB task observations and interviews about the implemented tasks. Based on these results, the theory on how quantitative and computational skills develop during data-driven programming course is presented. The theory consists of three main parts, that is, organization of code and data, construction of knowledge and understanding in data computation, and analysis to interpret data and draw conclusions. The authors conclude that the data-driven programming course improved students’ understanding of how computational tools can be used to process, analyze, and interpret data.

Relevance for Complex Systems Knowledge
The authors provide an approach which can be used to develop computational thinking skills. These skills are important in modelling complex systems, analyzing data, reasoning and identifying causal relationships. Although the study was carried with undergraduate biology majors, a similar approach can also be applied in other disciplines. The students had to visualize, interpret data and to explain various decisions made in programming, such as naming variables, labeling data, organizing the code. This approach enables students to understand the way data is stored in a dataset and improve understanding the logic flow in data analysis. However, authors determine that if computational tasks are related to only one type of programming, e.g. data visualization, students can misunderstand the concept of programming and relate it only to this programming area.

Point of Strength
The strength of the article is the study on how computational thinking can be developed by performing programming tasks in data-driven context. This scheme is based on reasoning and defining causal relationships between the analyzed data and the results.