ENHANCING COMPUTER PROGRAMMING FLUENCY THROUGH GAME PLAYING

Transcription

1 ENHANCING COMPUTER PROGRAMMING FLUENCY THROUGH GAME PLAYING Eshwar Bachu Department of Computing and Information Technology The University of the West Indies St. Augustine Campus, Trinidad & Tobago Abstract - In this paper we present an approach for learning and enhancing Pascal programming fluency through Game playing. This activity is intended to complement traditional teaching of concepts. The focus is on building program comprehension rather than program generation and enhancing programming fluency instead of programming literacy. In playing the game, students improve their ability to read and understand a program written in a specific language and to follow the logic in a program. They build speed in comprehension, as is required in learning any language. To win the game, students have to play certain steps repeatedly, using different strategies, and with time constraints. This repetitiveness reinforces learning and helps the students to become fluent with Pascal programming. Once they master the basic elements of a program, they will experience less frustration in coding solutions for more challenging problems. The preliminary results with students using this gaming activity are encouraging. Keywords - Pascal, programming, PascAl Shopper, computer games, fluency, literacy I. INTRODUCTION There has been an abundance of research carried out to investigate the teaching and learning of programming [2]; what emerges is that clearly computer programming is a difficult topic for many students. For novice programmers, a distinction can be made between programming comprehension (the ability to read and understand the outcomes of an existing piece of code) and generation (the ability to create a piece of code that achieves certain outcomes) [3]. In [4], the authors argue that a vital step toward being able to write programs is the capacity to read a piece of code and describe it. In [5], a comparison is made between IT fluency and IT literacy and the authors agree that fluency is the skill required to allow students to easily adapt as the world of IT evolves. Similarly, enhancing programming fluency Margaret Bernard Department of Computing and Information Technology The University of the West Indies St. Augustine Campus, Trinidad & Tobago Margaret.Bernard@sta.uwi.edu in students is vital to allow them to be more comfortable when entering CS1 and also enable them to be competitive in a rapidly evolving IT industry. In this paper, we present a computer programming game, PascAL Shopper, which is one of the activities in an e-course developed for teaching and learning computer programming [1]. The course is a two-year course designed for secondary school students (age group 14-16). It prepares them for the Caribbean Secondary Examination Certificate (CSEC) Information Technology examination administered by the Caribbean Examination Council (CXC). Programming has always posed a major challenge to these students; the PascAL Shopper game is designed to assist students with improving their programming skills in Pascal (Pascal is the recommended language of use for these students by CXC).The game provides students with Pascal code snippets which they are required to debug within a specified time; the game encourages the students to solve these problems through the added incentive of achieving the objective of the game. The game complements the other content in which computer programming concepts are taught using animation. Pascal Shopper was used by a group of secondary school students in an experimental study. We present the results of the evaluation and make some conclusions about our study. II. BACKGROUND Over the past four decades, tremendous research has been carried out into the teaching of programming to novice programmers, this has led to the creation of many programming languages and environments all aimed at making programming accessible to novices. Many of these innovations are documented and analyzed in [6] and [7]. More recently, new products such as Alice, Scratch and Greenfoot ([8], [9], [10]) which utilize the

2 technological advancements in animation and three dimensional modeling to appeal to the younger generation have been successful in engaging students in programming related activities. However, what remains clear is that none of these products have made a place into the classroom to be used as a compliment to teaching programming or has been widely categorized as successful. In [11], the author introduces a games based approach to teaching introductory computer programming. The author argues that the use of games can help with alleviating the difficulties which students face while programming and the common mistakes which they make. The concept of using games to help students identify common programming errors is documented in [12]. Also, from a student perspective, the use of a game also provides motivation and involvement which are crucial in the learning process. The aim of PascAL Shopper is to help students build program comprehension and programming fluency but before this can be accomplished, it is necessary that the students acquire programming literacy. This was primarily accomplished using the animation techniques presented by the authors in [1]. The students course of study included decomposing a problem, basic problem solving using selection, iteration, arrays and basic array search techniques. They also studied program design using pseudo code, flowcharts and other algorithm development techniques and finally the students would have completed introductory program implementation in Pascal. PascAL Shopper seeks to reinforce and compliment the traditional classroom delivery of this content. A. Gameplay III. DESIGN OF GAME PascAl Shopper is about a boy named Al (short for Algorithm) who has some shopping to do. Al has a list of six items which he must buy; these items can be found at three stores, a hardware, a supermarket or a fish market. Al is only allowed to ask for an item eight times throughout the game. If he exhausts his eight chances and does not collect all the items, he loses the game, this means that he must first ask for the item in the correct store (E.g. Asking for a hammer in the fish market instead of hardware) because if he asks for an item which is not sold in that store, a question is lost. When he does ask for the item in the correct store, the player will be presented with a multiple choice question based on Pascal programming which they must answer within a given time limit. The type of questions will be based on which level the player has chosen. There are four levels which can be chosen. The details of these levels can be seen in table 1. These four levels were included to allow the game to be used by all students, regardless of their current programming ability and it also allows the students to continuously challenge themselves. TABLE I. Description of the various levels Level Time Limit Question Description Easy Medium Hard Expert 15 seconds 40 seconds 40 seconds B. Game Features 120 seconds Tests General Knowledge of Pascal Programming Tests the ability to debug very simple, basic Pascal programs Tests the ability to debug more challenging Pascal programs Tests the ability to debug complex Pascal programs A major aspect of PascAl Shopper is that the game was customized to be more appealing to the target students. The main character in the game, Al, is a typical Caribbean character (Fig. 1). Also, the three main locations are decorated to resemble Caribbean scenery, for example, the fish market shown in Fig. 2 in which the fisherman is selling on the seashore reflects actual fisheries throughout the Caribbean. The three salespersons in the game were given varying physical descriptions to match the diversity of the Caribbean population and finally, the items which Al is required to collect throughout the game are also items commonly found throughout the Caribbean like red snapper, oysters and shrimp

3 C. Types of Programming Questions The technique used to evaluate the student s Pascal programming ability is a multiple choice question based on Pascal programming. In the easy level of the game, the player/student is presented simple questions which test general knowledge of Pascal programming. Examples of questions which can be asked are Which of the following is the assignment operator in Pascal? or Which of the following is used to obtain user input in Pascal?. However, in the medium, hard and expert levels, the player will be presented with an actual Pascal program or code snippet which they are required to debug (Fig. 3). Figure 1. PascAl Shopper main s character Al. A concerted effort was also made to ensure that the PascAl shopper game is very simple and easy to use while still challenging the students to complete the game. The game s screens were not crowded with unnecessary options and distractions, instead only the functionality required to the play the game was included (E.g. Asking for an item, returning to the Main Menu of the game and exiting the game) When the player gets a question incorrect, in addition to not receiving the item for which they asked, they are also not given the correct solution to the question. This was done to encourage students to play the game repeatedly even if they are not interested in winning the actual game. The questions which are presented to the players are randomly chosen from a question bank containing over sixty question, so while it is unlikely that players will be given the same question within a couple of attempts at the game, if they play the game continuously, they will get another chance at questions they may have previously gotten incorrect. In Fig. 3, we see that the player is asked to identify the error in the given code snippet. At the beginning of the snippet, the player is told what the code is supposed to do. At the bottom, the player is presented with four multiple options to choose the answer. A similar approach to the one described is used throughout the entire game. The game does not focus on the problem solving aspect of programming; instead the actual programming is tested. However it should be noted the problem solving programming aspects are still necessary for the user to be successful in the game. In Fig. 3, the player is already given the core of the solution and they are required to look for errors regarding the implementation of the solution in Pascal. All questions contain at most one error. Figure 2. Fish Market scene in PascAL Shopper

4 Figure 3. Screenshot showing question format in Pascal Shopper. D. Programming as a Foreign Language The idea behind using this approach can be compared to the learning of a foreign language. When learning a new language, reading and understanding words and small phrases are always the first aspects of the language which we learn. After mastering these, we move on to more complex sentence construction. Similarly, in PascAl Shopper, the player is presented with small Pascal code snippets which they are required to read to understand, they must determine what the outcome of the code is and what problem the code attempts to solve. Finally they must identify the error in the code if any exists. Using this approach, the students are forced to read and describe code and we believe that once the students have mastered this ability, the actual writing of Pascal programs will become easier for them. The need for students to become fluent in programming can also be compared to learning a new language, in that it is necessary that students repeat the words and phrases they learn until it becomes natural to them. Similarly, the use of Pascal Shopper will encourage students to repeatedly read and study pieces of code which will eventually lead to fluency since the students are motivated to play the game repeatedly until they win. In learning a foreign language, students are taught that when speaking to someone in the language, they should not translate the conversation into their native language; they should try and understand the foreign language directly. This is only possible when they have become fluent with the basic aspects of the language. Similarly, when the students have become fluent with basic programming constructs, their problem solving ability can improve because they will be able work through problem solving programmatically (i.e. as if they were writing code directly). This is important because many students are able to solve a problem but cannot translate the solution into program code. E. Implementation Since PascAL Shopper is an activity in an e- course (ITonLine), the game was implemented as Java applet to facilitate easy integration with the ITonLine software. The programming questions

5 which students are required to solve are stored in external text files and follow a specific format. The items on the shopping list as well as where these items can be bought were also stored in external text files. This makes it very easy to change the game at regular intervals to keep the students interested in playing. 5.8 which shows an average improvement of 11% after using Pascal Shopper. IV. EXPERIMENT A. Methodology The experimental study was performed on a group of twenty, form-five secondary school students. The experiment took place in the following manner 1) The students were given a short quiz which contained 10 multiple choice questions similar to those found in the medium level of PascAl Shopper. The duration of the quiz was fifteen minutes. 2) The students were then asked to play the medium level of PascAl Shopper. They were not told how many times to attempt the game. The students were not aware they would only be given forty minutes to play the game. 3) The students were given the same quiz they attempted before they played PascAl Shopper. The students did not know that they would be given a second quiz. The duration of the quiz was ten minutes. 4) Finally the students were asked to complete a questionnaire. It should be noted that the students had no prior knowledge of the experiment. Also, the students were not told about their performance in the first quiz before they played Pascal Shopper or attempted the second quiz. Figure 4. Figure 5. Pie chart representing students performance Line graph comparison of students score in two quizzes B. Experiment Results. The results of the experiments produced the following results. From Fig. 4, it is shown that a majority of the students performance increased in the second quiz from the first quiz after playing PascAL Shopper. Fig. 5 and Fig. 6 show the difference in the students performance between the two quizzes, they clearly show that most of the students made substantial improvements in their performances. The average mark in the first quiz was 4.7 (out of 10), the average mark in second quiz was Figure 6. Line graph representation of difference in quiz scores

6 90% of the students said that they would like to more activities like Pascal Shopper in their classroom environment. The questionnaire also asked the students to give recommendations about how the Pascal Shopper game can be improved and become more helpful to them. A range of responses were given, but the most common recommendation received was to add sporting gameplay instead of the shopping gameplay to the game. Figure 7. Pie chart showing the percentage of students who played between one and five games The breakdown of the number of games played by each student can be seen in Fig. 7. We see that 60% of all students played at least four games. Results of the experiment showed that 85% (17) of the students succeeded at winning the game and 59% (10) of them continued playing the game even after they had won. An analysis conducted on the results of the questionnaire completed by the students concluded the following; 85% of students said that they would have continued playing the game until they won. 85% of students thought that the game would help them to perform better on the second quiz. (The students were never told how they performed) Additionally, the students made a greater effort to understand and solve the questions in the second quiz than in the first one and they agreed that this was mainly because they felt more comfortable attempting the questions after they had played PascAl Shopper. Evaluating whether the game helped the students become very fluent is very difficult, but from the results of the quizzes and the fact that the students had a less time to complete the second quiz and their performance improved, we believe that continued use of the PascAl Shopper game will lead to the students becoming fluent in Pascal programming. V. EXTENSIBILITY AND FUTURE WORK Pascal, the programming language used in PascAl Shopper can be easily substituted for another language. The only requirement is substituting the external text files with questions and options from another programming language. The images of the characters in the game can also be easily substituted to allow the game to be used within other cultures. 95% of students agreed to try the other levels of the game 85% of students said they preferred playing Pascal Shopper as opposed to doing similar questions in a written format. 90% of students said that the game helped them understand basic Pascal programming better. 95% of students said that the game helped improve their skills in identifying common Pascal programming errors. Future work for PascAl Shopper includes providing a more efficient process through which teachers can add or update the questions in the game and the method of storing of the questions and answers will be changed to extensible markup language (XML). This will increase portability between PascAl Shopper and existing learning management systems, like Moodle which uses the MoodleXML format. VI. CONCLUSION The result of the experiment clearly shows that PascAl Shopper was useful in helping students

7 improve their basic comprehension of Pascal programs. The technique of using a game to evaluate and teach basic Pascal programming was also successful, the students were encouraged to play the game repeatedly, regardless of the fact that game comprised mainly of programming questions which these students would normally not be enthusiastic about doing. Continuous attempts at the game provided the students the opportunity to study Pascal code and become familiar with the common errors which they make and they would therefore be less likely to repeat these mistakes in their own programming. Additionally the student also develops the skill of being able to describe a piece of code as opposed to simply reading it. The time constraint on the questions was also very helpful because the students were required to think quickly about the code snippet and the problem it attempts to solve and this is a useful skill for an examination. The fluency gained by the students will also prove very useful in an examination as time is a major factor in every examination. The game also provides a chance for a student to understand the Pascal programming language since the questions in game presents a problem, and attempts to solve the problem using Pascal. The students can therefore learn from these questions while trying to solve them. Again, this aspect would be very useful for people who are new to programming and Pascal. We can conclude that games can be successfully utilized in teaching programming comprehension to students and helping them improve their programming fluency Bloom and SOLO Taxonomies, Proc. 8 th Autralasian Computing Education Conference, Australia, 52: , [5]. A.L. Werner, S. Campe, J. Denner, Middle school girls + games programming = information technology fluency. SIGITE '05 Proceedings of the 6th conference on Information technology education,2005. [6]. A. Pears, S. Seidman, L. Malmi, L. Mannila, E. Adams, J. Bennedsen, M. Devlin, J. Paterson, A survey of literature on the teaching of introductory programming, Proc. Working group reports on ITiCSE on Innovation and technology in computer science education, [7]. C. Kelleher, R. Pausch, Lowering the barriers to programming: A taxonomy of programming environments and languages for novice programmers, ACM Computing Surveys (CSUR),Volume 37, Issue 2, June [8]. I. Utting, S. Cooper, M. Kolling, J. Maloney, M. Resnick, Alice, Greenfoot, and Scratch A Discussion, ACM Transactions on Computing Education (TOCE), Volume 10, Issue 4, November [9]. B. Ward, D. Marghitu, T. Bell, L. Lambert, Teaching computer science concepts in Scratch and Alice, Journal of Computing Sciences in Colleges, Volume 26, Issue 2, December [10]. M. Kolling, The Greenfoot Programming Environment, ACM Transactions on Computing Education (TOCE), Volume 10, Issue 4, November [11]. R. Rajaravivarma, A games-based approach for teaching the introductory programming course, ACM SIGCSE Bulletin, Volume 37, Issue 4, December [12]. E.M. Nalaka S. Edirisinghe, Teaching Students to Identify Common Programming Errors using a game, Proc. 9th ACM SIGITE conference on Information technology education, REFERENCES [1]. A. Rudder, M. Bernard, S. Mohammed, Teaching programming using visualization, Proc. 6 th IASTED International Conference on Web-based Education, France, ACTA Press, [2]. J. Sheard, S. Simon, M. Hamilton, J. Lönnberg, Analysis of research into the teaching and learning of programming Proc. International Computing Education Research workshop, California, US, [3]. R. Watson, M. de Raadt, M. Toleman, Teaching and assessing programming strategies explicitly, Proc. 11th Australasian Computing Education Conference (ACE 2009), Wellington, New Zealand, [4]. J. Whalley, R. Lister, E. Thompson, T. Clear, P. Robins, P. Kumar, C. Prasad, An australasian study of reading and comprehension skills in novice programmers, using the

8