1. Introduction
Agenda 21 plan of UNESCO states that education is a way to solve potential problems threatening our future society [
1]. Teaching and learning for sustainable future program at UNESCO also declared that education “as one of the most powerful instruments for bringing about the changes required to achieve sustainable development [
2].” Specifically, regarding sustainable development goal on education by UNESCO, one of goals (SDG4) states that quality education and lifelong learning opportunities for all must be addressed [
3].
Information and Communication Technologies (ICT) as efficient pedagogical tools have the potential to respond and fulfill the SDG4 [
4]. For example, in Velázquez and Méndez’s study [
5], augmented reality and mobile devices have been confirmed to support student learning (quality education) and improve students’ skills in using emerging technologies (lifelong learning). In Deaconu et al.’s study [
6], ICT integrated in vocational education and training provided promising results in which students’ learning outcomes were improved (quality education) and specific skills were acquired for lifelong learning.
The current study examined the opportunities and difficulties in integrating ICT in education. In the study, one ICT (drone) was incorporated in an after school program at a public elementary school in Taiwan. Students participating in the curriculum were empowered to employ flying drones to foster their understanding of spatial visualization and sequencing (quality education) and to enhance their awareness of computational thinking (lifelong learning for future society).
5. Results and Discussion
Overall, the research design of the study only fitted in a specific learning scenario. Because of a small sample size, the following quantitative information only reported the phenomenon occurred at the after-school program and cannot be generalized into other learning contexts. In addition, the following qualitative information was used for interpreting unique learning patterns and corroborating the quantitative findings.
5.1. Instructional Effects of Drone Use
The results of descriptive statistics and
t tests are summarized in
Table 2 and
Table 3, respectively. These show that significant gains were identified in the sequencing (
t = 4.70,
p < 0.01) and spatial visualization (
t = 4.42,
p < 0.01) tests, with a major improvement (mean difference = 5.4) demonstrated for spatial visualization. Therefore, drone programming training could significantly enhance young students’ sequencing and spatial visualization skills development, and drone use had a particularly large learning effect on their spatial visualization.
5.2. Gender Effect
Although the study group was sex equivalent, the research team attempted to assess whether gender affected on students’ skills development. The
t test results for gender effect are reported in
Table 4 and
Table 5. Regardless of the measurement type, the statistical information indicated that gender as a potential variable did not influence students’ performance in sequencing (pretest:
t = 0.22,
p > 0.01; post-test:
t = 0.63,
p > 0.01) and spatial visualization (pretest:
t = 0,
p > 0.01; post-test:
t = 0,
p > 0.01). Thus, male students demonstrated the same learning patterns as female students throughout this educational experiment.
5.3. Drone Programming Patterns
Three major programming patterns were found through analysis of students’ work:
Pattern 1: No relationship between drone programming and skills development.
Removing the effects of the two pretests yielded the partial correlation analysis results reported in
Table 6. Overall, students’ programming work in the last stage of the learning progression model was not significantly related to their performance in the sequencing (r = 0.08,
p > 0.05) and spatial visualization (r = 0.24,
p > 0.05) posttests. Therefore, no relationship between students’ drone programming work and the development of their sequencing and spatial visualization skills was found.
Pattern 2: Avoidance of loop concepts.
Loop blocks in the programming language (Tynker) enabled students to reduce redundant blocks. However, most students preferred to use basic blocks to sequence their programming rather than use advanced loop blocks. For example, if a drone was designed to fly two times in a circle, students tended to repeat the same blocks twice; they would not employ the loop blocks to shorten the programming pattern. Although these two approaches yielded the same programming outcomes, students preferred the more inefficient programming design.
Pattern 3: Gender difference in drone movement.
When designing drone movements, male students demonstrated bolder learning patterns than their female counterparts. A flip block (
Figure 7) often appeared in male students’ programming work. Moreover, in the same learning scenario male students preferred wide flight movements, whereas female students focused on narrow flight paths (
Figure 8). The comparative frequency of flight movement patterns between male and female students was analyzed using a chi-square test, the results of which are summarized in
Table 7. The statistical findings showed a significant difference between male and female students for flip block use (
χ2 = 10,
p < 0.05) and flight pattern width (
χ2 = 10,
p < 0.05).
5.4. Students’ Learning Responses
Qualitative analysis of the informal interviews and class observation yielded the following five themes.
Theme 1: Learning enthusiasm
In weekly learning sessions, the students all demonstrated interest in learning about drone programming and motivation to engage in different types of learning activities. However, students’ desire to learn sometimes created a noisy environment in which they enthusiastically and loudly shared their work with instructors and peers. Under such circumstances, the instructors had to insist on classroom discipline to control student behavior. Students’ increased learning motivation could be attributed to the study’s innovative instructional design. Most students perceived the learning content of the experimental program as more attractive and enjoyable than conventional school learning. For example, one boy said, “I love to make something using technology. But my classes in school do not provide such opportunities. That’s why I was so excited about the program.”
Theme 2: Self-reported learning gains
Because these young students did not have programming experience, they perceived programming as the major educational benefit. Several students overestimated the complexity of programming and expressed a desire to learn programming tasks. For example, one girl said, “In the beginning, I was worried about my programming skills. But, to my surprise, the visual programming blocks were not very hard to understand. I began to love coding.” Or, as one boy stated: “Programming looked like building with physical blocks. It was not very hard and more fun than with traditional blocks. The most important thing was that it allowed me to generate something by combining different blocks.” The second self-reported learning gain was spatial training. Several students considered spatial visualization to be their learning weakness. However, with the assistance of drone use, they gradually gained knowledge of 3D space and could then mentally visualize flight direction and movement. For example, one boy said, “I often had difficulty identifying right or left. But I improved my spatial skills through drone programming design.” Or, as one girl stated: “When I began to design the drone flight path, I would visualize the direction in my mind. … When I got home, I even showed my mom and brothers how 3D space worked.”
Theme 3: Learning support from the instructor
During the first 2 weeks, perhaps because students were not familiar with programming language, they required constant learning support from the instructors. Under such circumstances, the three instructors were busy moving around the classroom to provide programming guidance, and this directly affected the pace of the lesson plan. As students gradually gained programming knowledge in subsequent weeks, a major problem with programming debugging appeared in the classroom. For example, students easily passed through the first two stages of the learning progression model (copy and tinker) but tended to fail to complete their programming work in the last stage (create). Several students frequently sought help because the flight movement did not match their programming design. Once the instructors explained the debugging principles, however, students were able to adopt them to solve problems.
Theme 4: Gender difference in programming block use
When asked why they had added more flip blocks in their programming work, all the male students responded that such flying tricks could generate creative programming. For example, one boy said, “Wasn’t it cool to flip the drone? It was so boring to have the drone just flying around.” Or, as another boy stated: “Flipping was such a dynamic movement. It added value to my work.” By contrast, some female students viewed flipping as a dangerous movement, whereas others thought that simpler programming blocks were adequate. For instance, one girl said, “I was afraid that using the flipping function I might hit something. I only used that block a couple of times.” Or, as another girl stated: “The other blocks already served my needs. I only wanted my drone to fly a specific movement, not do special tricks.”
Theme 5: Gender difference in flight patterns
Similar to the results in Theme 4, a gender difference was identified in flight patterns. When male students discussed their programming works during the interview process, they claimed to have done their programming work without thinking and gave no specific reasons. One boy said, “It is just what I did. I could not tell you the reason. Maybe it is my intuition.” Another boy said, “I have no idea. Probably all male students prefer a wide flying pattern.” Although female students similarly did not offer detailed reasons, they did hint at concerns regarding safety. One girl stated, “I realized that I made this pattern only when you pointed it out. Maybe I did not want my drone to hit something.” Another girl said, “I have no idea about my pattern. But the best reason for it might be that narrow flight would not affect other classmates’ learning.”
Theme 6: No preference for loop block use
The students were asked about using loop blocks. They were basically unanimous in stating that both methods achieved the same outcome. In their opinion, traditional block sequencing in programming work did not create an extra burden on them. Although the loop block reduced the numbers of blocks, they still preferred to build their blocks in repeated fashion. For example, one boy said, “I understood the function of the loop block. But I still liked using the traditional way.” Or, as one girl stated: “I knew that both ways would work. Even though I used a lot of blocks to build my project, the outcome was still accurate.”
5.5. Instructional Design Problems
Instructional design problems from the pilot and formal stages of the study are summarized as follows:
Issue 1: Protective measures
The blades of the drone posed a threat to student safety. During the pilot stage of the study, flying drones often accidently hit some of the schoolteachers and students, directly causing a severe skin wound. Therefore, creating drone management rules was necessary for student safety. For example, in the formal stage of the study when students were completing their programming work, they were required to obtain the instructors’ approval for their drone tests. Only one drone was allowed to take off at a given moment, and the other students were obliged to be aware of a drone flying in their surroundings. Although some students may have to wait some time for their drone’s departure, preventive measures can create a safe learning environment.
Issue 2: Space selection
Originally, the targeted learning setting was a normal-sized classroom. However, after several tests during the pilot stage of the study, the physical size (i.e., height and width) of the traditional classroom was found to limit the drone’s flight potential. A spacious room with a higher ceiling such as an assembly hall or a small sports stadium is a perfect location for instruction in flying drones. In addition, to avoid wind effects, electric fans or air conditioners in the selected locations should be turned off when students are ready for drone testing. For example, in the formal stage of this study, electric fans often affected the accuracy of drone flight movements.
Issue 3: Bluetooth interference
The lightweight drones used in this study were connected to students’ tablet computers via Bluetooth communications. Because only two drones were used for instruction preparation in the pilot stage of the study, Bluetooth interference problems did not occur at that time. However, in the formal stage of the study when 10 students were using tablet computers simultaneously, Bluetooth signals from the computers could interfere with each other. For example, one student’s tablet computer would unintentionally connect to a classmate’s drone. To solve this problem, students were requested to verify that the serial number labeled on the drone matched the one shown on their tablet before testing their drones.
Issue 4: Programming check
During the formal stage of the study, students performed drone testing without instructors reviewing their programming work. However, a landing problem often arose because several students had forgotten to add a landing block, and thus their drones continued to float in the air. For a safe landing, the students were required to terminate the programming language (Tynker) to stop the flying drones. Therefore, to avoid potential safety problems the instructors were advised to conduct a programming check before the drones took off.
Issue 5: Power supply
Testing in the pilot stage of the study showed that the battery life of lightweight drones only permitted 20 min of flight (in standby mode battery life could be longer), and therefore one battery for each student was not sufficient for practice. To facilitate learning, each drone was equipped with three backup batteries during the formal stage of the study. Moreover, although a drone management rule was created to ensure student safety, accidents such as unexpected wall impacts severely damaged the blades, thus terminating the drone’s flight functions. The instructors had to prepare a reserve supply of blades.
5.6. Overall Discussion
The statistical findings demonstrated a significant improvement in young student’s spatial visualization and sequencing skills over the 6-week educational experiment. Drone programming could thus have instructional effects on students’ spatial visualization and sequencing skills. Given that drone programming enabled students to gain knowledge of visual block programming, the results of this study supported those of Kazakoff et al. [
26], who reported that using a Lego robot with visual block programming significantly enhanced young children’s sequencing skills. Moreover, a mean comparison indicated that the improvement in spatial visualization learning surpassed that of sequencing, perhaps because the effect of the instructors’ intervention had a greater effect on spatial visualization. These findings might also be attributed to drone flight movement design, which forced students to constantly employ spatial thinking to understand 3D geographical locations [
25]. Several students also indicated similar learning gains in the qualitative findings.
Other findings of this study indicated that gender as a potential variable did not have an effect on students’ spatial visualization and sequencing; boys and girls demonstrated a similar performance for each measurement. However, a gender difference was identified in drone programming patterns. Compared with their female counterparts, male students tended to use bolder patterns in their programming design (i.e., flip blocks and wide flying). This was fully explained in the qualitative findings where boys stated that they were more likely to adventurously use unique programming blocks and design wide flight movements, whereas girls performed their programming tasks more conservatively, perhaps because of safety concerns. Although the study participants were elementary school students, these results reflected the findings of another study that identified a gender difference in programming patterns in a college computer course [
43].
Students’ block use preferences showed a specific programming pattern, namely, less use of loop blocks. Most students preferred to use basic blocks to complete their programming work even though loop blocks enabled them to reduce redundant blocks and make their programming more efficient. The reason for this was identified in the qualitative results, in which most of students stated that they viewed the loop concept as optional because they could use other blocks to achieve the same purpose. In another study, lack of loop block use was suggested as a possible common programming style [
44] in young children. Another programming pattern was that students’ drone movement design was unrelated to their spatial visualization and sequencing performance. The reason for this could be that the programming design process enabled them to practice spatial thinking and programming sequences, resulting in major learning improvements in these two measurements. Students’ programming work in the last stage of the learning progression model exhibited only outcomes rather than process.
Students participating in the drone programming program all showed an enthusiasm for learning, echoing the findings reported by Hussey [
15] and Wakefield [
16] on drone use by young children. Another reason for students’ desire to learn was perhaps a positive comparison between this program’s innovative curriculum and conventional school learning [
39]. Drone programming offered attractive and enjoyable content, whereas normal classroom instruction did not provide such learning opportunities for them. In addition to learning motivation, learning support is a primary task in instructional settings. When facing difficulties in the learning process, young children require instant learning support provided by instructors. Guidance in programming and debugging principles might thus have served as learning scaffolding [
45] that constantly decreased students’ cognitive learning load.
On completion of the two-phase research model, five instructional design problems emerged: protective measures, space selection, Bluetooth interference, programming check, and power supply. Relevant solutions were proposed to achieve instructional effectiveness [
40]. Because two of these problems, protective measures and programming check, related to students’ personal safety, they must be emphasized and treated as instructional priorities [
12]. Other problems such as Bluetooth interference and power supply directly influenced students’ learning. Without adequate preparation and support, the pace of instruction and time for programming practice might be greatly compromised. Regarding space selection, an ample space, particularly one with a high ceiling, could fully develop the instructional potential of drone use, depending on the available school infrastructure.