A collection of visualization-based tutorials on graph algorithms (http://plonedev.uwosh.edu/jhave/algorithms-for-learners/graphs) was used in data structures classes at both the Helsinki University of Technology and at the University of Wisconsin - Oshkosh. The course at HUT was taught by Archie Korhonen in Spring 2008, and the course at UWO was taught by Tom Naps in Spring 2009. Korhonen’s course had over 100 students in it, and Naps’s course had only 12. Apart from these differences in class sizes the visualization-based tutorials used by both groups were nearly the same. The tutorials covered depth-first search, breadth-first search, Dijkstra’s shortest path algorithm, and the Prim and Kruskal minimal spanning tree algorithms. For each algorithm, students used two separate visualizations delivered by the JHAVE AV viewing environment — one presented in the Matrix visualization plug-in developed at HUT and one presented in the GAIGS visualization plug-in developed at UWO. Perhaps the key difference in these two visualization plug-in for JHAVE is the way in which they have the user engage with the algorithm being studied. Matrix has the student simulate the execution of the algorithm by manipulating the data structure (via mouse interaction with the nodes in the structure) at each successive step. Hence it uses what the engagement taxonomy would call the changing mode of engagement. The GAIGS visualization engine has the student answer pop-up questions about the algorithm, which would exemplify the responding level of engagement. What was difference between the reactions of the two groups of students? Before working through these tutorials on graph algorithms, the students at HUT had previously only used the Matrix viewer to study other algorithms whereas the students at UWO had previously only used the GAIGS viewer. We surveyed the students about their experience in using both visualization plug-ins. Perhaps not surprisingly, each group indicated a greater comfort level in using the visualization plug-in with which they were familiar from previous course work. In some cases, they expressed minor frustration that having to learn how to interact with a new visualization plug-in made them concentrate less on actually learning the algorithm. Although no statistically significant conclusions can be drawn from our survey results, they indicate that, when using visualizations throughout a course, providing students with the same GUI for delivering those visualizations may be the ideal. It will maximize the time they spend thinking about the algorithm as opposed to struggling with learning a new GUI. Should you wish to try the tutorials at the above site, if you don’t have an account at jhave.org, you will not be able to use the visualizations in “quiz-for-real mode” in which your interactions with the visualization are logged to the JHAVE database. Instead follow the directions in the tutorials about how to run the visualization in “practice mode”.