5 DANCING SORTING ALGORITHMS (STUDY 3)
In order to practice loop instructions and arrays, the majority of computer programming curriculums include sorting algorithms: bubble sort, insertion sort, selection sort, etc. In our third research (Kátai & Tóth, 2010), we analysed how dance can be integrated in the teaching-learning process of the first two sorting algorithms (bubble and insertion sort).
We invited students who like dancing to collaborate in our project. They played the roles of the numbers from the sequence to be sorted. Each of them wore the corresponding number on their dress. The accompanying music pieces were composed on the basis of Michael Flatley’s music.
Analysing the above-listed algorithms, we realized that the choreographies have to include two basic elements: 1. comparing two numbers, 2. swapping two numbers. Moreover, bubble sort and insertion sort algorithms can be implemented
Figure 5.1. Bubble sort dance performance; dancers 7–9 have already reached their final positions. During the fourth traverse,
dancers 5 and 6 perform a changing operation.
Figure 5.2. Insert-sort dance performance. The main character of the sixth act of the algorithm, dancer 7, has just begun her
insertion process.
in such ways that only neighbouring elements need to be considered. Consulting with the eurhythmics teacher of the faculty, we found that a proper dance step for comparison operations would be that the corresponding two dancers simply turn face-to-face to each other. Regarding the swapping operations, we chose different dance steps adequately to the corresponding dance music. During the performances, dancers that are not in focus are dancing alone, on the spot. In order to make it perceptible that each traverse of the sequence moves at least one number to its final position (in the case of the bubble sort algorithm), dancers who have reached their final place change their dancing style (see Figure 5.1).
With respect to insertion sort algorithm, at the beginning of the performance, all dancers dance with their back turned to the audience. Each act of the algorithm starts with the next dancer turning to the front. Then, this dancer, as a kind of main character of the current act, inserts him-/herself in the sequence of the already sorted dancers by a comparing-and-swapping sequence (see Figure 5.2).
Once these dance performances had been videotaped, we added further graphical elements to the records in order to emphasize that dancer numbers are stored in an array and to highlight the dance couple in the focus. We also included role-playing in the multisensory sorting algorithms lessons. Differences between the stage choreographies and the classroom scenarios were as follows:
dance steps were replaced by moves, and the dance music was removed.
5.1 Software tool and suggested syllabus
The multimedia software tool we created helps students analyse the studied sorting algorithms. Some of its characteristics are the following:
– It can be used as both live and self-paced e-learning tool.
– It makes possible backsteps in the interactive animation process. If stu-dents miss out on certain details during the stepwise simulation of the algorithm, they can ask the software to step back and repeat the problem-atic operation.
During sorting algorithm lessons, we applied the following syllabus:
– Teacher presents the strategy the algorithm applies and simulates it on a given number sequence.
– Teacher discusses with students about implementation aspects.
– Students view the dance performance of the algorithm. The teacher, as a narrator, makes references to implementation aspects.
– Students are invited to role-play the algorithm.
– Students simulate the algorithm using the multimedia application referred above.
– Students implement the algorithm.
57
5.2 THE EXPERIMENT
– As homework, students are asked to work out choreography for selection sort algorithm.
Discussing with the students, they revealed that some of them had the fol-lowing learning experience:
– After they had attended the explanation and followed the simulation on the number sequence, they thought that they understood how the algo-rithm works.
– Watching the dance performance, they realized that their understanding of the algorithm had been refined.
– During the role-playing, they observed further details.
– Working at home on selection sort choreography, they understood further subtleties of sorting algorithms.
5.2 The experiment
In order to investigate empirically if dancing and role-playing indeed en-hance teaching-learning sorting algorithms, we performed the following expe-rience. Thirty-eight first-year undergraduate students of Sapientia Hungarian University of Transylvania (Târgu-Mureş, Romania) were involved in the exper-iment. All students started to learn C programming language at the beginning of the 2008/2009 school year.
Before the experiment, we organized a pre-test to form two statistically equivalent groups (experimental/control). In the pre-test, we tested students’
knowledge in applying loop instructions. The averages of the pre-test points of the students assigned to the experimental and control groups were 7.06 and 6.4 respectively (according to the 1–10 marking scale in use in Romania). Comparing these performances (independent sample t-test), we did not receive significant differences (p = 0.35 > 0.05).
During the experimental period, the 19 members of the experimental group were taught according to the new multi-sensory method. (Since we had proposed to investigate the impact of dance and role-playing on programming education, we excluded from the syllabus the use of the software tool.) In the control group, the 19 students were taught according to classic methods, without making any effort to involve senses in the teaching-learning process.
Once the experiment ended, students’ knowledge in sorting algorithms was tested (post-test). In the following, we listed the problems the students had to solve:
– Simulate the two algorithms on a given sequence.
– Implement an insertion sort/bubble sort strategy for sorting in decreasing order.
– Improve the teacher-presented bubble sort algorithm.
– Implement the selection sort strategy.
The averages of the points received for the post-test done by the students of the experimental and control groups were 6.73 and 4.47 respectively. Comparing these values with the independent sample t-test, we found that the difference between them is significant (p = 0.003 < 0.05) (favouring the experimental group).
In conclusion, it can be stated that the results of this experiment support our expectation that dancing and role-playing can improve students’ skills in analysing, designing, and implementing sorting algorithms.
59
5.2 THE EXPERIMENT