Research On Visualization Of Program Error-finding Process Based On Eye Movement

Carrying out behavioral analysis of students’ programming and debugging ability is very important for cultivating students’ reading and comprehension skills and way of thinking.In recent years,it has gradually become a hot topic in the field of programming education.Compared with traditional evaluation methods,eye tracking data can reflect the distribution and shift of students’ visual attention during the programming process,and provide a more direct insight into the thinking process.However,multiple dimensions of information such as time series eye movement data,code logic characteristics,and student task intentions are intertwined,making pure quantitative analysis showing increasingly greater limitations,and new analysis methods are urgently needed.This thesis takes the webpage reading code error-finding test process as the research scene,fully records the student’s interactive actions and eye tracking process,and carries out a routine quantitative analysis of visual attention behavior.Aiming at the limitation of quantitative analysis that it is difficult to reflect the complete details of complex multidimensional data,this thesis proposes a visual design plan for two types of issues that are usually concerned in teaching activities.First of all,a design scheme is proposed to distinguish the difference of student code tracking ability into color band sequence.In order to characterize students’ tracking ability during the code reading process,based on the integrated processing of the code area and the eye movement jump sequence,the eye movement jump to the tracking state of the code execution logic is color-marked and dimensioned.In order to display the change and distribution information of different states,a visualization solution for the tracking state is obtained.Secondly,a design scheme is proposed to represent the process of the error finding hit as a convergent and oscillating process centered on the problem-solving task.Aiming at the visual attention characteristics of the same thinking task as the goal in the problem-solving process,the physical meaning is reorganized and weighted in the temporal space and the code line space,and the distribution of visual attention and the continuous migration process are mapped into discrete chromaticity blocks.In this way,a visualized scheme of the chessboard chart aiming at the process of the error finding hit by a single person is obtained.By measuring the range of the convergence phenomenon in the hit process,and extracting and superimposing it,a visualization solution of the convergence graph for the commonality of the group performance can be obtained.Based on the "C Language Programming" course,the analysis scheme was applied to the actual data of the case where 19 undergraduate students read the wrong code three times,and the robustness of the sample size was verified on the data set of 31 students.Experimental results show that the above scheme can achieve the purpose of portraying the commonality of individual tracking status differences and group attention performance.The corresponding visualization results can also be used as a reference basis for other aspects of the error-finding process.It has an auxiliary function for teacher teaching and guiding significance for the evaluation of learning effect.