Proceedings of the Annual Meeting of the Cognitive Science Society

Existing research has drawn parallels from the comprehension of text to the comprehension of source code. In this study, we attempt to develop this analogy by positing and testing a notion of code cohesion, analogous to text cohesion. We also attempt to extend a known effect in text comprehension research, the reverse cohesion effect, to code contexts. Our findings provide some corroboration for code cohesion, but fail to find robust evidence for a reverse cohesion effect. This reinforces similarities between text and code comprehension but also suggests that everyday comprehension processes of code and text might differ in meaningful ways.