An Experience of Introducing Primary School Children to Programming using Ozobots
Nina Körber, Lisa Bailey, Gordon Fraser, Barbara Sabitzer, Marina Rottenhofer
AAn Experience of Introducing Primary School Children toProgramming using Ozobots
Nina Körber
University of PassauPassau, Germany
Lisa Bailey
University of PassauPassau, Germany
Gordon Fraser
University of PassauPassau, Germany
Barbara Sabitzer
Johannes KeplerUniversity LinzLinz, Austria
Marina Rottenhofer
Johannes KeplerUniversity LinzLinz, Austria
Abstract
Algorithmic thinking is a central concept in the context of compu-tational thinking, and it is commonly taught by computer program-ming. A recent trend is to introduce basic programming conceptsalready very early on at primary school level. There are, however,several challenges in teaching programming at this level: Schoolsand teachers are often neither equipped nor trained appropriately,and the best way to move from initial “unplugged” activities to cre-ating programs on a computer are still a matter of open debate. Inthis paper, we describe our experience of a small INTERREG-projectaiming at supporting local primary schools in introducing childrento programming concepts using Ozobot robots. These robots havetwo distinct advantages: First, they can be programmed with andwithout computers, thus helping the transition from unpluggedprogramming to programming with a computer. Second, they aresmall and easy to transport, even when used together with tabletcomputers. Although we learned in our outreach events that the useof Ozobots is not without challenges, our overall experience is posi-tive and can hopefully support others in setting up first encounterswith programming at primary schools.
Keywords
OzoBots, Programming Education, Primary School Programming
Computational thinking [16] is an important aspect of education,and programming is a primary method in order to teach it [7, 8].Establishing an elementary understanding of algorithmic thinkingearly on is important to successfully achieve computational think-ing, but also to counter common preconceptions against program-ming and computer science that influence the gender imbalancelater on [9, 11, 13, 14]. Programming is also increasingly used asan instrument to teach other topics. The best methods for teachingprogramming early on, however, are still being explored.A common approach to introducing children to programmingconcepts is by using “unplugged” activities that require no comput-ers [1, 4], for example by letting the children impersonate robotsexecuting sequences of instructions. A challenging next step isto transfer concepts learned with unplugged programming to pro-gramming with computers. Physical computing with programmabledevices such as robots is a possible intermediary step, as they havea tangible, real-world impact through their actions [3, 18]. However,this requires schools to have access to appropriate hardware, andteachers need to have the confidence and knowledge to managehardware devices and to support children in using them.While computational thinking is already integrated into the cur-riculum in many countries from the earliest stages on, German and
Figure 1: The Ozobot Evo robot: Small, easy to handle, with asoft plastic cover that makes it sturdy. The colourful fingersof the child in the picture are a common result of the pen-and-paper mode of interaction.
Austrian primary schools currently do not touch upon the topic. Inan endeavour to help local primary schools with the introduction ofprogramming concepts, we therefore visited eight primary schoolsfrom October 2018 to October 2019. During these school visits, weused the Ozobot Evo programmable robots in two-hour workshops.Ozobots are small programmable robots for which initial experiencereports are generally positive [5, 15, 17]. The Ozobots have severaladvantages: They bridge the gap between unplugged programmingand physical programming, as they can be programmed with penand paper as well as on a tablet or desktop computer; they are verysmall and portable, and can easily be moved from one school to an-other; the programming environment is very flexible and the levelof difficulty can be adapted from pre-reading stage up to complexprogramming concepts.In this paper, we describe the workshop format we developedfor our outreach activity, and share our experience of interactingwith children, teachers, and Ozobots. Overall, the children wereenthusiastic as one would expect, both when interacting with penand paper as well as with tablet computers. The Ozobots proved tobe better suited for their pen-and-paper mode than programmingwith a tablet computer. Figure 1 shows an Ozobot Evo robot. The robot can make sounds,has multiple lights that can flash in different colours, and it canmove. It has sensors to detect the colour of the surface it is drivingon, and proximity sensors facing to the front and back. https://ozobot.com/ a r X i v : . [ c s . C Y ] A ug ina Körber, Lisa Bailey, Gordon Fraser, Barbara Sabitzer, and Marina Rottenhofer Figure 2: Ozobot robot moving along a line with embeddedcolour codes. In this example, green-blue-red increases thecounter by one (the symmetric colour sequence red-blue-green decreases the counter).Figure 3: The OzoBlockly programming environment in aweb browser: Categories of blocks are shown on the left, andclicking on a category opens a drawer of blocks available. Atlevel 1 (“Pre-Reader”) only a few blocks are shown, but manymore blocks are available at increasing levels.
There are two main ways of interacting with an Ozobot: In theline-following mode, the robot follows coloured lines it senses withits bottom sensors. While doing so, it changes its lights based on thecolour of the line it is following. Lines may contain special coloursequences, which represent commands (e.g., turn left, turn right, gofast, dance, etc.) Figure 2 shows an example of the robot following aline while increasing a counter. The main categories of commandsare related to speed (from snail dose to nitro boost), special moves(spinning, zigzagging, etc.), directional movement (turning, jump-ing lines, u-turns, etc.), counters (e.g., counting colours, countingcrossings, increasing/decreasing counters), timers, and commandsto indicate success or completion of a task .Ozobots can also be programmed via the OzoBlockly app (Fig-ure 3). The programming environment follows the standard block-based approach with a toolbox of different categories of blocks onthe left hand side of the screen. OzoBlockly supports five differentlevels, where in the easiest level blocks are purely graphical withicons and only a few numbers instead of text, so that programmingcan be done even at pre-reading stages. The basic commands are https://play.ozobot.com/print/guides/ozobot-ozocodes-reference.pdf, last accessedMarch 15, 2020 https://ozoblockly.com/, last accessed March 15, 2020 reminiscent of the Logo [12] programming approach, and mainlycontrol the movement of the robots. At the highest level, the pro-gramming language supports even complex constructs such as func-tions with return values. OzoBlockly programs can be transferredto the robots by “flashing”, which means holding the robot facedown on the screen while the program is transferred by a coloursequence. When using a tablet rather than a desktop computer,the Ozobot Evo can also be controlled via an app and a bluetoothconnection. The app contains not only OzoBlockly, but also variousfunctionalities such as a remote control. The bluetooth connectivityis one of the central differences between the Ozobot Evo robot andthe cheaper variant, the Ozobot Bit: The Ozobot app to programrobots using a tablet computer can currently only be used with theEvo, while programming the Bit always requires flashing programs,and thus a web browser for operation . Our outreach activity was organised in the form of two-hourworkshops at the primary schools, targeting children in grades 3–5.We visited eight schools between October 2018 and October 2019.Each workshop was supervised by 2–4 project participants, andthe class sizes were around 20–25 students (sometimes coveringtwo classes at once). In this section, we describe the sequence ofactivities conducted at the workshops. These activities are basedon a selection of teaching material found in the large collectionavailable online as well as our own ideas. After a brief introduction of the organisers, we discussed withthe children what a robot is, what robots they know, and how theylook like. Typically, they expected robots to be large and humanoid.Then, we handed Ozobots to them (one robot for two children). Asinitial activity, we asked the children to find out what equipmentand capabilities the robot has, and how to turn it on. After a while,we let the children tell us what they had discovered and wrote downthe correct suggestions on the blackboard. This activity proved tobe helpful in order to set realistic expectations, for example, therobots do not have elastic springs and therefore are not able to jump(which was particularly useful to understand for when the childrenwere later introduced to “line jumping” commands, which referonly to switching from lines to follow rather than physical jumping).Ozobots have a single button that is not entirely obvious and tookthe kids a while to find. They discovered the wheels and the USBcharging port quickly, and once they had managed to turn on therobot, they discovered the lights and sounds the robot can produce.The proximity and colour sensors are not obvious and needed someexplaining; this immediately lead to the second activity.
To explain how the robot uses the colour sensor, we handed outthe first worksheet , which consists of a black line with gaps init. The children received coloured pens and were asked to fill thegaps and observe how the robots behave. There were two learning There is a separate app for the Bit which is focused on the line-following mode. https://ozobot.com/educate/lessons, last accessed March 16, 2020 https://storage.googleapis.com/ozobot-lesson-library/3-5-basic-training-color-codes/3-5-Basic-Training-Color-Codes-full-version.pdf, p. 24, last accessed March 16,2020 n Experience of Introducing Primary School Children to Programming using Ozobots objectives: First, the children were supposed to discover that theOzobot follows lines, and while doing so, the lights glow in thecolour that the colour sensor detects (white instead of black, as onlyexception). Second, the children got a feeling of how thick the lineshave to be so that the robot can detect them. The first worksheetwas followed by a second one that also consists of lines with gaps,but this time the gaps are intentionally located at corners and bendsin order to practice drawing appropriate lines the robot can follow.Children that completed the task quicker than others were allowedto use the time for extending the lines or drawing their own lineson the back side of the paper. Before starting the third activity, we discussed the experienceof the second activity with the students to make sure everyonehad noticed how the colour sensor works. This lead to the thirdactivity, which is concerned with how the colour sensors can beused to control the robots. To introduce the children to this concept,the third handout contains a circular line with two gaps, and twocolour codes that the children need to draw into the gaps. Once thechildren had had some time to try out these codes, we discussedwith them how the two codes caused the robots to behave. Eachcode consists of a sequence of three colours; in this first example,both codes are symmetric, so that the robot behaves identical re-gardless of the direction in which it traverses the code. With thefourth handout , the children again received a line with two gapsand two corresponding codes, but this time only one code is sym-metric, while the second one has a different effect depending onthe direction of traversal (in one direction it slows the robot down,in the other it accelerates it). Again, we discussed this with thechildren after they had experimented with the handout. The fifth handout consists of two basically equal paths, eachcontaining a crossing which leads to three different colours; inthe second path, there is a gap for inserting a command. As firstactivity, we tasked the children to conduct an experiment wherethey start their robot ten times on the first path, and keep track ofhow often the robot moves to each of the colours. Once completed,we collected the data on the blackboard, to see which colour wastargeted most of the times (to simplify the counting, we only askedeach child or pair of children which colour was visited most fre-quently). Since the Ozobot makes a random choice at a crossingunless instructed otherwise, the result is different every time. Afterexplaining that the robot chooses randomly, we handed out tableswith the main colour codes that the Ozobot can handle, and turnedto the second path on the handout. After explaining the differentcommands on the command table, we chose an arbitrary colour(usually the one that was least frequently visited during the previ-ous experiment) and then tasked the children to select and draw acommand that will make the robot go to that colour every time. At this point, the children had an understanding of the capabil-ities and the behaviour of the robots and they knew how to use Ibid., p. 25 Ibid., p. 26 Ibid., p. 27 Ibid., p. 28 colour codes as well as which commands the robot can understand.We thus handed out a challenge, chosen from the many availabletasks on the Ozobot website. In these challenges, there typically isa scenario (e.g., the road from home to school ) with a clear objec-tive (e.g., get the robot from home to school) and many obstaclesthat require the use of colour codes (e.g., dead ends, road works,etc.) Solving such a challenge sometimes took a while, and it wasuseful to be prepared to handle mistakes (either by handing out anew sheet, or by using white stickers to cover incorrect commands).Children who completed the task earlier than others were allowedto create their own racetrack on the back of the handout. The next activity was the transition from pen-and-paper com-mands to sending commands via programming. This activity wasdone without robots (we tried to use this time for recharging therobots a bit). We discussed the general concept of sequences withthe children, and introduced the commands in the OzoBlockly for-mat using large paper printouts of the easiest OzoBlockly level,which consists of graphical blocks. Using these printouts, we fo-cused on the computational thinking concept of sequences [2]: Wepinned sequences of commands on the black board, and then askedthe children to execute the programs (which typically consisted ofwalking, turning, and making happy/sad faces). While one childexecuted the program, the other children were tasked to judge ifthe execution was correct or not. Finally, we demonstrated theOzoBlockly app on a tablet computer, showing how the same com-mands can be arranged in sequences there. In the final activity, each pair of children was given a tabletcomputer (and a recharged Ozobot, if necessary). We explainedhow to connect the tablet computer to the robot via bluetooth,and how to use the OzoBlockly app. Then, we gave the childrena handout consisting of the first coding challenge . We used theOzobot challenge where the robot has to go from the start positionto its room in its house, and pre-loaded the OzoBlockly app with anincorrect implementation of the sequence of instructions, leadingto the robot erroneously ending up in the forest. The task for thechildren was to correct the program so that the robot successfullywalks to its home. Consequently, the main computational thinkingconcept reinforced by this activity is again sequences , as well as thecomputational thinking practice [2] of testing and debugging . Oncethe task was completed, the children were given the chance to getcreative at producing an exciting program. Over the course of theworkshops, we refined this activity to the task of programming theOzobot to perform a dance. In addition to the school workshops we organised an eventthemed as “Programming Olympic Games”, where we invited fourof the classes previously visited during workshops to the Universityof Passau. Within two days, the students were able to demonstrate Ibid., p. 29 https://storage.googleapis.com/ozobot-lesson-library/ozoblockly-training-k-1/ozoblockly-training-k-1.pdf, pp. 14-20, last accessed March 16, 2020 Ibid., p. 13 ina Körber, Lisa Bailey, Gordon Fraser, Barbara Sabitzer, and Marina Rottenhofer (a) Incorrect program (starting point) (b) Expected correct solution(c) A workshop participant conducting the programming activity
Figure 4: The first OzoBlockly programming task we used inour workshops: The task is to let the robot move home safely.The starting point is an incorrect program that erroneouslymoves the robots to the forest. The correct solution requiresreordering the rotation instructions as well as replacing theemotion command at the end.Figure 5: The Dancing on Ice activity: Two teams of childreneach programmed a Ozobot robot to perform a joined danceroutine. One of the Ozobots is decorated with a unicorn skin. their skills acquired in the workshops and competed against eachother in four “Olympic disciplines”.The prime discipline was the
Ozobot on Ice activity. The childrenwere given 20 minutes to rehearse a choreography for two Ozobotsperforming a pair dance on ice. As can be seen in Figure 5, thechildren used OzoBlockly for this. Some of the criteria for evaluation included the synchronicity of the two Ozobos and whether theymoved within the ice rink or surpassed the line.We further included two disciplines using BeeBots, which areprogrammed with buttons with arrows for each direction and ago, cancel and stop button, located on the back of the robots. TheBeeBots move in 15cm-steps and usually, they are used with BeeBotmats which display a grid to facilitate the navigation. The userinterface only supports sequences of commands to be programmed,which matches what all children had learned during the workshops.The children had no problems in applying their Ozobot experienceto the BeeBots. We used BeeBots for two activities: In the
RallyeDakar with BeeBots children had to navigate the BeeBots alonga racetrack using their sequencing skills, by entering the rightcommands into the robot so that the BeeBot reaches the finish lineof the racetrack in one pass. In
BeeBot Bowling no BeeBot mat wasused and so the children had to correctly estimate distances. Thegoal of this game was to enter as many commands that are neededfor the BeeBot to knock over the cones (empty plastic bottles).
Independently of the project underlying this experience report,we organise a CoderDojo at the University of Passau at regularintervals. In the two-hour programming club young programmersaged 8 to 18 voluntarily come together to learn programming andhave fun. The children usually work on their individual projects bythemselves, with mentors always available to help them with theirproblems. Some of the children wanted to work with the Ozobots.The workshop format we had developed works well with classeswhere everyone works on the same project, but in our experienceit did not work well in the CoderDojo setting. Introducing Ozobotsrequired one of the instructors to be present all the time, as we hadnot prepared any tutorials in German and the more complex blocksrequire a level of English skills which not all of the children had.Some of them ended up having fun with the remote control insteadof programming the robots. With one of the classes we tried the “Modeling animal habits”lesson . After introducing the concept of point-counter colourcodes using one line without crossings as depicted in Figure 2,we gave the children the handout consisting of a map of crossinglines with empty spaces for colour codes. The task was to usepoint-counter colour codes and let the Ozobot count from five tozero points, taking into account the non-orthogonal design of thecrossings and which route the Ozobot will most likely take. Thechildren understood the task well and filled the map with suitablecolour codes. After counting from five to zero, the robot is supposedto stop and blink red. However, not in all cases it did, even thoughthe solution of the children was correct and the robot had passed therequired amount of colour codes in the right order. As this lack ofinstant positive feedback disappointed the children and explainingthe point counter colour codes was time-consuming, we did notrepeat this exercise in successive workshops. https://coderdojo.com/ https://portal.ozobot.com/lessons/detail/modeling-animal-habits, last accessedMarch 16, 2020 n Experience of Introducing Primary School Children to Programming using Ozobots Pen-and-Paper Drawing.
The children are fascinated with draw-ing creative patterns for the robots to follow with the Ozobot mark-ers (which can be replaced with regular markers once they are outof colour). When producing handouts or using other paper, makesure not to use standard printer paper but thicker paper, otherwisethe desks will be coloured.
Pen-and-Paper Tasks.
When given tasks where children have tofill in colour codes, mistakes will be made. It is important to eitherhave white stickers to allow corrections, or to have spare copies ofthe handouts. In the worst case, a reasonable workaround is to drawdetour lines to allow the Ozobots to bypass erroneous solutions.
Programming with OzoBlockly.
Programming the robots withOzoBlockly worked well as the icons are self-explanatory and thestudents intuitively understood how to drag and drop blocks. Oncethe complexity and length of programs increased, it turned outthat deleting blocks was less self-explanatory and often needed ex-plicit instruction. In Scratch [10], hat blocks can be used to specifyat which point in time a sequence of blocks, a script , is executed.OzoBlockly does not have such event handler blocks and in combi-nation with scripts which were not deleted, this led to unexpectedrobot behaviour, as the order and number of scripts executed isunclear when more than one script is present in the editor.
Understanding the Repeatable Nature of Programs.
Some ofthe children did not intuitively understand the repeatable nature ofprograms. In Activity 7, the task was to get the robot from a fixedstarting point to the house. Some of the children tried to use theeditor as remote control by running several different programs untileventually the robot reached the goal from wherever it stoppedafter execution of the previous program. When moving the robotback to the fixed starting point and running the current programagain, the robot did not reach the house. After explicitly hinting atthe goal to write one coherent, repeatable program with the robotstarting from the fixed starting point, the children usually got itright quickly. When using BeeBots at the Olympic Games, the samemisunderstanding happened, but again the children understoodwell after additional explanations.
Correct Programs Appearing Erroneous.
In Activity 7, some-times the execution of a program did not succeed even though theprogram was correct. The reason was that in consecutive execu-tions the robot usually did not have the exact same direction atthe starting point, therefore ending at a slightly different positionat every execution. However, as soon as the children had under-stood, they could differentiate between the usual small deviationpresent even for correct programs and programs which were indeederroneous, leading to the robot obviously missing the end point.
Complexity of Programs.
Using the pen-and-paper mode madeit easy for the children to understand how to use colour codes toprogram the robots. As in Activity 4 they already had to thinkabout the temporal interplay of commands, the transition fromdrawing to programming with commands seemed quite easy forthem. However, due to the limited amount of time, the complexity ofthe programs the children created with the OzoBlockly editor waslimited. Most of the time, we did not introduce loops or conditionals.In the cases where we did, the children had created very long scripts with repeating blocks and were happy and eager to learn about abetter solution for their code.
Labeling Robots.
It is helpful to name all the robots, and to addstickers with their names to them. Not only for matching tabletcomputers with robots and connecting via bluetooth the namelabels are true life savers, the children also identify much strongerwith their robots if they have names. We used two different sets ofrobots, one set with robots being labeled with numbers, and theother set with robots having real names such as “Fred”, “Lena” or“Bobo”, and in every class the favourites were the ones with names.
Remote Control.
In some of the classes we demonstrated that therobots can be controlled using the remote control functionality inthe OzoBlockly app. As it is easier and apparently more fun to usethe joystick than to create programs, some children started playingwith the robots rather than programming them. We recommend tohide the remote control functionality and hope that the children donot discover it themselves. If they do so, one has to try to convincethe children not to use it. For example, you can tell them that theymiss out on practicing to be a cool programmer when only playingwith a toy that even toddlers can handle.
Bluetooth Kills Battery Life.
While Ozobots can easily survive90 minutes following lines, the bluetooth connection seems to drainbatteries extremely quickly. We found it necessary to rechargerobots before switching from drawing mode to programming mode,and even then do the robots not last very long. Thus, we also recom-mend to turn off Ozobots during phases of discussion or instruction,and to plan for breaks where the robots can be recharged. Ideally,having spare robots to replace robots with dead battery is helpful.
Preparing Programs for Students.
The erroneous programs usedfor Activity 7 had to be created manually on every tablet computer,which was a time-consuming process. We strongly recommend totry the newly launched Ozobot classroom platform as it claims tosolve this problem. We have not tried it yet because it was launchedafter the end of our outreach activities. As an alternative, one canlet the students copy the erroneous code, which is also printed onthe handout, into the OzoBlockly editor themselves. This way, thetablets do not have to be prepared in advance, and the studentscan practice drag-and-dropping the blocks without having to thinkabout the code they want to write. Pairing Robots and Tablets.
It is officially recommended thateach tablet computer is matched with exactly one robot. We foundthis not to be a viable solution since robots run out of battery tooquickly when connected via bluetooth. However, when connectingdifferent robots and tablet computers, it frequently happens thatthe app refuses to control the robot until it has been “updated” –even if the robot is already completely up to date. Performing suchan update requires a working internet connection, which may notbe available during outreach activities. Also, both duration andsuccess of these updates were rather non-deterministic, which ledto some robots not being usable with tablets at all. The Ozobotsupport was not able to recommend an alternative at the time wecontacted them. Another problem which occurred repeatedly wasthat an already connected robot was not available in the editor ofthe tablet computer. Our solution to cope with this bug was to exitthe editor and then open it again. https://ozobot.com/educate/classroom, last accessed March 16, 2020 ina Körber, Lisa Bailey, Gordon Fraser, Barbara Sabitzer, and Marina Rottenhofer Checking Which Codes Are Needed.
The colour code referencesheet we used in our workshops (the English basic version ) doesnot include the code for “pause”, even though this is needed forActivity 5. Also, the codes for the counters needed in the “Modelinganimal habits” lesson are not included on the basic sheet. We choseto draw the missing codes on the blackboard. As an alternative, onecan use the more detailed sheet , but it might be easier for thestudents to orient themselves when having less codes to look at.A German version of the colour code reference sheet has beenreleased after our activities. In any case, we recommend checkingwhich codes are supposed to be used for the activities planned, sothat one can provide the students with the information they need. Language Barriers.
The OzoBlockly app supports several lan-guages, but the set of languages available is very limited. One ofthe reasons why we used the pre-reader level for the introductionwas indeed that German is not a supported language. Although wecontacted the Ozobot company, their response was that additionallanguages will be added only once there has been sufficient demand.At the time of this writing, apparently this has not been the casefor German, but we hope that this will change in due time.
Preconceptions against Programming.
Overall, most of the chil-dren were excited to work with robots, and for both boys and girls,being good at programming was something worth striving for. Girlswho did a good job at programming were happy to receive positivefeedback and be called good programmers. We did not recogniseany gender-specific preconceptions against programming or com-puter science, and boys and girls were particularly fascinated byone of the robots which we decorated with a unicorn skin (Figure 5).
Introducing children to programming concepts at primary schoollevel aims at fostering computational thinking skills and overcom-ing preconceptions and gender imbalance. In order to support thisendeavour, we organised programming workshops at local schools.Due to their size and functionality the Ozobot robots turned out tobe well suited for this task. Despite some challenges such as batterylife, on the whole our experience was overwhelmingly positive.Our workshops so far were limited to two hours, and thus onlyallowed us to teach basic programming concepts; in most cases wecovered no more than sequences. A particular feature of Ozobotrobots is their dual mode of operation in a pen-and-paper modeas well as block-based programming mode. In our experience thepen-and-paper mode is a great way to introduce children to theabilities of the robots, and to build up enthusiasm for robots in gen-eral. While the relevance of the initial activities in pen-and-papermode with respect to computer science learning taxonomies [6] canclearly be argued, it remains unclear whether the concrete skillslearned during this mode of operation support the subsequent learn-ing of more advanced concepts related to programming.Therefore, as a next step we would like to explore the use ofOzobot robots when continuing with introducing further conceptsand gradually progressing learners to more advanced programming. https://files.ozobot.com/stem-education/ozobot-ozocodes-reference.pdf, p. 1, lastaccessed March 19, 2020 Ibid., p. 2 https://drive.google.com/file/d/19UBAsQ2_SEXBaZJgfbIxD8gH-lgTYkfh/view, lastaccessed March 19, 2020 With each of the five levels in the OzoBlockly app the complexityincreases, and the top level seems equally well suited for olderlearners. Due to the possibility of viewing a JavaScript representa-tion of OzoBlockly programs, it is even conceivable to use Ozobotsthroughout the learning process, up to text-based programming.
Acknowledgements
This work was supported by Kleinprojekt NB-23 “Grenzüber-schreitende Förderung der Informatikbildung” funded by theINTERREG-Programm Österreich-Bayern 2014–2020.
References [1] Tim Bell, Jason Alexander, Isaac Freeman, and Mick Grimley. 2009. Computerscience unplugged: School students doing real computing without computers.
The NZ Journal of Applied Computing and Information Techn.
13, 1 (2009), 20–29.[2] Karen Brennan and Mitchel Resnick. 2012. New frameworks for studying andassessing the development of computational thinking. In
Proceedings of the 2012annual meeting of the American educational research association , Vol. 1. 25.[3] G Barbara Demo, Giovanni Marcianò, and Simonetta Siega. 2008. Concreteprogramming: Using small robots in primary schools. In . IEEE, 301–302.[4] Hylke H Faber, Menno DM Wierdsma, Richard P Doornbos, Jan Salvador van derVen, and Kevin de Vette. 2017. Teaching computational thinking to primaryschool students via unplugged programming lessons.
Journal of the EuropeanTeacher Education Network
12 (2017), 13–24.[5] Rostislav Fojtik. 2017. The Ozobot and education of programming.
New Trendsand Issues Proceedings on Humanities and Social Sciences
4, 5 (2017).[6] Ursula Fuller, Colin G Johnson, Tuukka Ahoniemi, Diana Cukierman, IsidoroHernán-Losada, Jana Jackova, Essi Lahtinen, Tracy L Lewis, Donna McGeeThompson, Charles Riedesel, et al. 2007. Developing a computer science-specificlearning taxonomy.
ACM SIGCSE Bulletin
39, 4 (2007), 152–170.[7] Katharina Geldreich, Mike Talbot, and Peter Hubwieser. 2018. Off to new shores:preparing primary school teachers for teaching algorithmics and programming.In
Proc. 13th Workshop in Primary and Secondary Computing Education . 1–6.[8] Sze Yee Lye and Joyce Hwee Ling Koh. 2014. Review on teaching and learning ofcomputational thinking through programming: What is next for K-12?
Computersin Human Behavior
41 (2014), 51–61.[9] Helen M Madill, Rachel G Campbell, Dallas M Cullen, Margaret-Ann Armour,Albert A Einsiedel, Anna-Lisa Ciccocioppo, Jody Sherman, Leonard L Stewin,Stanley Varnhagen, T Craig Montgomerie, et al. 2007. Developing career com-mitment in STEM-related fields: myth versus reality.
Women and Minorities inScience, Technology, Engineering and Mathematics (2007), 210.[10] John Maloney, Mitchel Resnick, Natalie Rusk, Brian Silverman, and Evelyn East-mond. 2010. The Scratch Programming Language and Environment.
ACMTransactions on Computing Education (TOCE)
10 (11 2010), 16.[11] Susan Staffin Metz. 2007. Attracting the engineers of 2020 today.
Women andminorities in science, technology, engineering, and mathematics: Upping the numbers (2007), 184–209.[12] Seymour Papert. 1980.
Mindstorms: Children, Computers, and Powerful Ideas .Basic Books, Inc., USA.[13] Claude M Steele. 1997. A threat in the air: How stereotypes shape intellectualidentity and performance.
American psychologist
52, 6 (1997), 613.[14] Amanda Sullivan and Marina Umaschi Bers. 2013. Gender differences in kinder-garteners’ robotics and programming achievement.
International journal oftechnology and design education
23, 3 (2013), 691–702.[15] Diane van der Linde, Nicole van der Aar, and Joke Voogt. 2018. Best of TheNetherlands: How children use computational thinking skills when they solve aproblem using the Ozobot. In
EdMedia+ Innovate Learning . Association for theAdvancement of Computing in Education (AACE), 2151–2157.[16] Jeannette M Wing. 2006. Computational thinking.
Commun. ACM
49, 3 (2006),33–35.[17] Martin Žáček and Pavel Smolka. 2019. Development of Computational Thinking:Student Motivation Using Ozobot. In
Proceedings of the 2019 3rd InternationalConference on Education and E-Learning . 36–40.[18] Goran Zaharija, Saša Mladenović, and Ivica Boljat. 2013. Introducing basic pro-gramming concepts to elementary school children.