The software engineering industry is increasingly aware of the role and value of neurodiverse engineers within the workforce. One motivation is the alignment between skills needed for software development and the processing strengths of individuals with autistic spectrum conditions. One aspect of neurodiversity is dyslexia, typically presenting in individuals through a range of reading deficiencies. In this paper we build on recent work which has sought to investigate if programmers with dyslexia read program code in a way which is different from programmers without dyslexia. The particular focus of this analysis is the nature of saccadic movement and patterns of linearity when reading code. A study is presented in which the eye gaze of 28 programmers (14 with dyslexia and 14 without) was recorded using an eye tracking device while reading and understanding three on-screen Java programs. Using insights from the wider dyslexia literature, hypotheses are formulated to reflect the expected saccadic gaze behaviour of programmers with dyslexia. A range of existing metrics for linearity of program reading are adapted and used for statistical analysis of the data. Results are consistent with recent work elsewhere and indicate that programmers with dyslexia do not exhibit patterns of linearity significantly different from the control group. Non-linear gaze is shown to be approximately 40% of all saccadic movement. Some preliminary insights are offered based on the data available, suggesting that the extent of non-linear reading when comprehending program code might complement the processing and problem solving style of the programmer with dyslexia.
|Title of host publication||Evaluation and Assessment in Software Engineering (EASE 2020), April 15–17, 2020, Trondheim, Norway. ACM, New York,|
|Publisher||Association for Computing Machinery|
|Number of pages||9|
|Publication status||Published - Apr 2020|
McChesney, I., & Bond, R. (2020). Observations on the linear order of program code reading patterns in programmers with dyslexia. In Evaluation and Assessment in Software Engineering (EASE 2020), April 15–17, 2020, Trondheim, Norway. ACM, New York, (pp. 81-89). Association for Computing Machinery. https://doi.org/10.1145/3383219.3383228