Efficient evaluation strategies are essential when reviewing computer code for potential reuse. Previous researchers have examined the factors that influence these assessments. However, researchers have yet to empiric...
详细信息
ISBN:
(纸本)9780998133171
Efficient evaluation strategies are essential when reviewing computer code for potential reuse. Previous researchers have examined the factors that influence these assessments. However, researchers have yet to empirically demonstrate the direct influence of the specific factors that affect visual/cognitive effort, which can be inferred through eye tracking metrics. Programmers were recruited to complete a Java code review task, providing evaluations for a code file's trustworthiness and reusability after various errors had been introduced to the file's source reputation, readability, and organization. Analyses of the eye-tracking data revealed increases in fixation counts and durations for manipulated code. An exploratory analysis of areas containing readability and organization errors revealed misuses of case and misuses of declarations garnered the most attention from participants relative to the rest of the code. Implications of the current study extend to recommendations for writing code that is easily reusable by decreasing the visual effort needed for code review.
Computer code has entered our society in contexts ranging from medical to manufacturing settings. The current study expanded previous literature by examining the effects of three between-subject factors (i.e., reputat...
详细信息
ISBN:
(数字)9783030490652
ISBN:
(纸本)9783030490645;9783030490652
Computer code has entered our society in contexts ranging from medical to manufacturing settings. The current study expanded previous literature by examining the effects of three between-subject factors (i.e., reputation, organization, and readability) on various trust-related outcomes. Participants (N = 54) were computer programmers recruited from Amazon's Mechanical Turk (MTurk). We used a 2 (reputable or non-reputable source) x 3 (high, medium, or low organization) x 3 (high, medium, or low readability) between-subjects design to examine how the independent variables interact to predict the trustworthiness perceptions of the code. The results show that programmers perceive code differently when coming from reputable sources. Thus, it is important to highlight whether or not any open source code comes from a reputable source and make this information readily available to programmers. Another trend we found is that programmers tend to prefer conspicuously high or low organization, particularly when readability is low. Thus, a medium level of organization could obfuscate the goals of the original programmer, which may undermine the programmer's intent and reduce codetrustworthiness.
暂无评论