Tue 8 Aug 2023 09:00 - 09:25 - Code Comprehension

Background and Context. Cognition, in all its forms, is inferential; inferential reasoning underlies processes like decision-making, problem-solving, argumentation and text comprehension. In the Psycholinguistics domain, it is argued that the study of text comprehension- the way that a coherent representation or a mental model of the text is constructed - is a study of inference generation. In programming education, code comprehension has been considered a milestone for students’ progression. An important goal of code comprehension tasks is for the readers to build coherent representations or mental models of the code.

Objective-Hypothesis. A common activity during which code comprehension is required is studying a solved programming problem (SPP). There is a threefold comprehension process that takes place during SPP tasks: comprehension at the problem level, comprehension at the program level, and their relationship. These three processes guide the comprehension and the construction of a coherent mental representation of the whole example; drawing from Psycholinguistics, we hypothesize that differences in the way that students construct the mental model of a given SPP may be associated with differences in inference generation during these three processes.

Method. Through a mixed research design that brings together a theoretical-conceptual study and an empirical one, we first develop an inferential model for SPP based on scholarly research in Psycholinguistics and Programming Education. The inferential model is then tested and further refined through a comparative case study during which we identify and compare the kind of inferences a proficient, an average, and a struggling first-year undergraduate student made during the comprehension of an SPP task after the end of the first semester.

Findings. The results of our study demonstrate that the more advanced students employed specific inferential strategies during the three SPP comprehension processes: Global Backward Explanations and Forward Predictions, Local Backward Explanations and Forward Predictions, Associations and Paraphrasing, all of which guided their SPP comprehension and successful transfer to an isomorphic task.

Implications. The paper provides the first evidence towards a) the importance of inference generation during the study of SPP tasks; b) the kind of inferences made during successful SPP comprehension; c) the development of new research directions and theoretical insights and concepts in the area of SPP comprehension; and d) the development of teaching practice that elicits inferential strategic reading comprehension in order to explore and guide students’ understandings before actual problem-solving.

Maria Kallia is an assistant professor in the School of Computing Science (Centre for Computing Science Education) at the University of Glasgow.

Tue 8 Aug

Displayed time zone: Central Time (US & Canada) change

09:00 - 10:15
Code ComprehensionResearch Papers

Session Chair: Sally Fincher

The search for meaning: Inferential Strategic Reading Comprehension in Programming
Research Papers
Maria Kallia University of Glasgow
Examples of Unsuccessful Use of Code Comprehension Strategies: A Resource for Developing Code Comprehension Pedagogy
Research Papers
Colleen M. Lewis University of Illinois at Urbana-Champaign
Chronicles of Exploration: Examining the Materiality of Computational Artifacts
Research Papers
Michael J. Johnson Georgia Institute of Technology, Francisco Castro New York University, Betsy Disalvo Georgia Institute of Technology, Kayla DesPortes New York University