The Influence of Environmental Odors on Student Programmers During Code Comprehension and Code Writing

Optimizing productivity is a longstanding goal in Software Engineering (SE). From models of development to IDE tools, researchers have investigated the factors that help or hinder programmers. One research avenue has examined nonfunctional factors like interruptions and environmental noise. Odor is a pervasive but unstudied factor in SE, with implications for workplace design, management, and assistive technology. To test odor's influence on programming tasks, we conducted a rigorously controlled study using eye tracking to measure participants’ cognitive processes during code writing and comprehension questions. Thirty student programmers were split into three groups exposed to different odors: Lemon (pleasant), Neutral, and Fish (unpleasant). Unaware of the odor manipulation, participants answered code comprehension, code writing, and prose comprehension questions while their gaze was recorded. We found that Lemon may have helped participants finish code comprehension questions more quickly (25.9% faster than Fish, \(p<0.05\) ; 7% faster than Neutral, \(p=0.1\) ). For code writing, both Fish and Lemon appeared to affect performance, with significantly higher visual effort than Neutral (longer fixation durations, \(p<0.05\) ). Our study offers a unique perspective on environmental factors in SE, suggesting pleasant odor may aid code comprehension, while no odor may be best for code writing.