Summary
The main aim of the master thesis was a eye-tracking evaluation of salience and interaktive work in visual programming languages.
The first part of the thesis is about visual salience. The aim was a preparation and realization testing of salience by the method of eye-tracking, and then a comparison of calculated visual salience in software OGAMA and results of eye-tracking testing.
Four visual programming languages for eye-tracking testing were selected. They were Model Maker (ERDAS Imagine 2013), Macro Modeler (Idrisi 17.0 The Selva Edition), ModelBuilder (ArcGIS 10.1 for Desktop Advanced) and Workflow Designer (Autodesk AutoCAD Map 3D 2014). Five experiments for eye-tracking testing of salience were created. Individual test was prepared for four selected VPL components. The number of stimuli was twelve in each experiment except the last one. This experiment contains the data flow diagrams of all four VPL components and the number of stimuli was forty-eight. Last experiment was for respondents with no experience with VPL components. The type of stimuli was image. The images change after two seconds.
Other step was to calculate a visual salience in software OGAMA. The results are saliency maps with the order of fixations. The results of eye-tracking testing of salience and calculated visual salience was compared. The comparision was commented verbally. After evaluation and comparison, it was found that respondents were influenced the way of reading data flow diagrams.
In the second part of the thesis a evaluation of interactive work in selected visual programming language was performed. The testing of interactive work by using method of eye-tracking was held in VPL component ModelBuilder. Two experiments for eye-tracing testing were created. The number of stimuli was five in each experiment. The type of stimuli was screen recording. In first testing the number of participants was ten and in second testing the number of respondents was twelve. The eye-tracking evaluation of interactive work was very time-consuming. It was necessary to watch all records from individual participants. The overview tables of operations were created. For selected operations were written time information into the tables. Then individual tasks were commented verbally.
