Seeing confusion through a new lens: on the impact of atoms of confusion on novices’ code comprehension

Studying the Impact of Continuous Delivery Adoption on Atoms of Confusion Rate in Open-Source Projects

“…As said before, Atoms of Confusion are easily identifiable and indivisible patterns of code capable of making the developer misunderstand the code they're in and how it executes. While the original study and list of Atoms of Confusion was based on the C programming language [8], different programming languages have different Atoms of Confusion candidates, and there are already studies that proposed their own lists of AoC candidates for specific programming languages [3,5,11,18]. By definition, AoCe can be transformed into other patterns of code that function in the same way and which are less likely to be confusing, making the use of AoC unnecessary.…”

“…Gopstein et al [8] introduced the concept of Atoms of Confusion as the smallest piece of code that can cause confusion in developers, making them misunderstand what the code actually does, which can lead to mistakes during development and when doing tasks. This work also focused on AoC in the context of the C programming language, while other works studied the AoC in different programming language contexts, such as Castor [3], with Swift; Langhout and Aniche [11], with Java; Torres et al [18], with JavaScript; and [5], with Python. Other works also study the relationship of AoC with different metrics, such as Bogachenkova et al [2] which analyzed the possible relation between AoC, code review, and pull requests.…”