| Abstract (english) | This thesis focuses on creating a conceptual framework for the gamification of Croatian borndigital dictionaries. The conceptual framework is illustrated by examples from Mrežnik, a borndigital dictionary that is being created at the Institute of Croatian Language and Linguistics within the research project Croatian Web Dictionary – Mrežnik financed by the Croatian Science Foundation. The Mrežnik project aims at creating a free, monolingual, easily searchable, hypertextual, corpus-based, online dictionary of the Croatian standard language with 10,000 entries. These entries contain definitions, subentries, collocations, examples of usage, normative and pragmatic advice, idioms, references to other entries, and external links. Mrežnik is based on the Croatian Web Repository corpus and the hrWaC – Croatian web corpus. In addition to these sources, all other available print and web sources are taken into account in writing definitions and providing examples and meanings. Mrežnik is created using the TshwaneLex dictionary compilation software that allows the creation of the desired structure and fields for dictionary entries. Mrežnik consists of three modules. In the first phase of the project, the module for adult native speakers of Croatian will have 10,000 entries, the module for school-children will have 3,000 entries, and the module for non-native speakers will have 1,000 entries. Games based on the dictionary content for all three models are developed within the project. The conceptual framework of gamification of born-digital dictionaries takes into account the process of developing, testing, and implementing these games within the dictionary. At the end of the Mrežnik project, the online dictionary with its games will be publically available on the web site rjecnik.hr. Gamification is a process in which game elements are implemented in non-gamelike situations to make them more fun. Gamification is used in many areas such as education and fitness. Many learning applications such as CodinGame for learning computer languages and Duolingo for foreign language learning use games (mostly quizzes) and gamification elements such as scoring, leaderboards, badges, and skill trees. A lot of research using games for learning purposes has been conducted but it does not focus on gamifying the dictionary content and testing it for learning purposes. Technology for creating games has progressed and now there are a lot of web services such as H5P, Quizlet, EclipseCrossword, etc. and free game codes with no copyright restrictions which are available on sites like CodePen and GitHub for creating educational games. Before creating dictionary games for the Mrežnik dictionary, the types of educational games and gamification elements used on websites of lexicographic publications had to be identified. Hence, the first research question in this thesis is: ● RQ1: Which gamification elements are present in Croatian and foreign lexicographic e-publications?To answer this research question, an analysis of 181 web dictionaries and 71 encyclopedias was conducted. The websites were found on the Wikipedia list of lexicographic publications and academic search engines RefSeek and iSEEK. The research showed that only 26 web dictionaries and 10 web encyclopedias contain games or gamification elements. Games identified on these sites are quizzes, drag-n-drop games, puzzles, crosswords, fill in the blanks games, listening games (games for spelling of heard words), offline materials for games, games for finding words, memory, typing games, hangman, and unique games. Within each of these games and on websites of online dictionaries and encyclopedias these gamification elements were identified: scoring, levels, time limit, leaderboards, avatar, badges or other award systems, and story with quests. Dictionaries, unlike encyclopedias, do not have any unique games. A story with a quest is the least present gamification element in lexicographic games. To get more data about educational game and gamification elements in addition to the previous research question the research was expanded with the next research question: ● RQ2: Which gamification elements are present on the library, archives, and museum websites? To answer that question 179 museum, 254 library, and 21 archive websites were analyzed. It was determined that only 18 libraries, 11 museums, and 2 archives contain games, links to games, or gamification elements. . Same as in dictionaries and encyclopedias on archive, library, and museum websites quizzes are the game type that occurs most frequently and scoring is the gamification element which occurs most often. Stories and quests are present more often on these websites because they are appropriate for presenting cultural and historic content as it is easier to tell a story through games. On the bases of this analysis and the identification of game types and game elements from educational websites, the first demo versions of games for Mrežnik were created. The demo versions of these games were presented to schoolchildren and non-native speakers learning Croatian as a foreign language. Their reactions were noted, analyzed, and taken into consideration for further development of games. Games for Mrežnik were created using available free technology that had no copyright restriction. Most of the games were developed through available codes on Codpen and GitHub websites, and some games were developed through the H5P platform. The demo versions of the games were published online through GitHub and GitLab repositories. Currently, many different game types were developed such as quizzes, crosswords, typing games, fill in the blanks, drag-n-drop, memory, etc. and they implement mentioned gamification elements such as scoring, leaderboards, levels, time limits, and badges. The finished games organized by categories will be published on the address: rjecnik.hr/igre. To test the effectiveness of these games, after getting the permission from the Ethics Committee for this research, the research was conducted on two groups of students who learn Croatian in Croaticum – Centre for Croatian as a Second and Foreign Language. The research was conducted in online classes during the pandemic of the COVID-19 disease (from 6th April to 22nd April 2020). The experimental group (9 students) used games published on GitHub during their class: https://borna12.github.io/croaticum/. The control group (12 students) had the same lesson but did not use educational games in class. To compare their progress, the same vocabulary test was given to students in both classes. At the beginning of the first class, they wrote a pretest to analyze and compare their initial knowledge of the vocabulary that is to be learned in the lesson. The posttest, which is the same as the pretest but the questions are in a different order, was given to the students of both groups at the end of that class to see their progress after the class. After two weeks they wrote a delayed posttest, which was again the same as the previous two tests, to check their long-term vocabulary retention. The results of all three tests were used to test the hypotheses: Ha0: Students who learned Croatian as a foreign language and are going to use gamified materials did not achieve significantly different results in the pretest than students who are not going to use gamified materials. ● Ha1: Students who learned Croatian as a foreign language and are going to use gamified materials achieve significantly different results in the pretest than students who are not going to use gamified materials. ● Hb0: Students who learned Croatian as a foreign language and used gamified materials do not achieve significantly better results in Croatian vocabulary tests than students who have not used gamified materials. ● Hb1: Students who learned Croatian as a foreign language and used gamified materials achieve significantly better results in Croatian vocabulary tests than students who have not used gamified materials. ● Hc0: Students who learned Croatian as a foreign language and used gamified materials achieve significantly lower or equal progress in Croatian vocabulary tests compared to students who have not used gamified materials. ● Hc1: Students who learned Croatian as a foreign language and used gamified materials achieve significantly better progress in Croatian vocabulary tests compared to students who have not use gamified materials. ● Hd0: Students of Croatian as a foreign language after using gamified materials achieve significantly lower or equal results in Croatian vocabulary tests. ● Hd1: Students of Croatian as a foreign language after using gamified materials achieve significantly better results in Croatian vocabulary tests. ● He0: Students of Croatian as a foreign language who like electronic games did not achieve significantly different results in Croatian vocabulary tests from students who do not like electronic games or are neutral to them. ● He1: Students who learn Croatian as a foreign language and like electronic games achieve significantly different results in Croatian vocabulary tests from students that do not like electronic games or are neutral to them. ● Hf0: There is no significant correlation between students' results in Croatian vocabulary tests and their self-evaluation of their current knowledge of the Croatian language. ● Hf1: There is a significant correlation between students' results in Croatian vocabulary tests and their self-evaluation of their current knowledge of the Croatian language. Hypothesis Hf was tested with the Spearman correlation, hypotheses Hb and Hc were tested using a one-way Mann-Whitney U-test, hypotheses Hd was tested using Wilcoxon signed-rank test, and hypotheses Ha and He were tested using the two-way Mann-Whitney U-test. The calculation for all these tests was done using the programming language R. The power of each test was calculated using the programming language R or program G*Power to measure the power of the result given by the statistical test based on sample (n) and effect size (d). Based on the results of statistical and power tests alternative hypothesis in Hd was accepted that students achieve significantly better results after using gamified materials. For Hb only in the analysis of the delayed posttest, which analyses long-term memory, the null hypothesis was accepted that students who use gamified materials after two weeks achieve equal or worst results than students who were not using gamified materials. For other hypotheses, nullhypothesis would be accepted if the power of the test was higher so no valid conclusion concerning these hypotheses could be made. However, based on descriptive statistics for Hc in the first posttest results and delayed posttest results, we can see that students from the experimental group have improved after the pretest, more than the control group. Students from both groups also completed a short questionnaire after the first posttest. Most students (17) mention that they use a web dictionary for learning Croatian. Eleven students (six from the experimental group, five from the control group) said that they like playing electronic games, nine students are neutral to them and only one student does not like to play electronic games. Students from the experimental group also answered a question about the games used during their class. Eight out of nine students enjoyed learning through games. Those students agreed that these games are good, simple, and fun for learning the Croatian language. Two students commented that some games are maybe too simple for advanced learners and that more descriptive words and examples of word usage could be implemented. The results of the conducted research implicate that games have the potential to help students in learning Croatian as a foreign (other) language. The methodology for conducted research can also be used for future researche with a larger sample which could give more statistically significant results. Based on observation during class, research results, and students’ feedback current games developed for the Mrežnik project were modified to allow more difficulty adjustments and have more examples of using a certain word or explaining a certain language rule. More adjectives were added to motivate the player. The next hypothesis tests the possibility to create a conceptual framework for the gamification of a web dictionary based on previously identified gamification elements. To prove the hypothesis, we must take into consideration which educational game types and gamification elements are present on the dictionary and encyclopedia websites as well as on the websites of other educational institutions. We must also take into consideration how these games are created, tested, and implemented within born-digital dictionaries and how can we check whether they are successful with users. Based on that, seven steps of the conceptual framework of gamification of born-digital dictionaries were designed: 1. deciding on the educational contents and goals of the game; 2. designing game types and deciding which gamification elements to use; 3. game development; 4. game testing and updating; 5. game publishing; 6. game promoting; 7. monitoring user satisfaction. The steps in the designed conceptual framework are mostly linear so when one step finishes the next step begins. However, game creators can return from step seven to step four to fix bugs or mistakes or update the game content based on user feedback. When creating a conceptual framework for gamifying e-dictionaries, the first step consists of choosing dictionary content that the creator of games wants to gamify and selecting the targeted group for which the gamified content is developed (e.g. elementary school children, non-native speakers, university students). In Mrežnik, games based on dictionary content were categorized as spelling games, phonological games, morphological games, syntactic games, discourse games, lexical games, word formation games, and games for learning the elements of the Croatian culture. Within each of these categories, games for the three modules of Mrežnik (adult native speakers, school-children, and non-native speakers) were developed. The starting point for the second step is to decide what type of games will be developed for the selected dictionary contents and for the targeted players selected in step one. When the game type is decided on the process of choosing which gamification elements can be implemented within the selected game types (timelimit, levels, badges, etc.) and which assignments created from dictionary content will be incorporated into the game. For this step, it is important to know which game types for educational games exist and what gamification elements can be used. This was analyzed in this thesis within the first two research questions. On the example of Mrežnik, many game types are presented with their gamification elements which are used for gamifying all categorized dictionary content. When games and game elements have been selected the development of games can start (step three). For game development, first the adequate technology that can allow creating games based on decisions made for previous steps has to be chosen. Sometimes this technology, code, or program for developing games has to be tested to see if it allows game creation based on the concept for the desired game. If testing proves so, the selected technology can be used for final game development. After the first version of the game is created, the process of testing the game, fixing bugs, and potentially updating game content can begin (step four). Game testing can be done by the game developer himself, but it is also recommended to have other users, that represent the targeted audience for the game, test the game, and give their feedback. If the game, after a few tests, is playable, has no major problems, and covers everything planed in the concept, it can be published. In the case of Mrežnik, games were created using available free technology and demo versions of the games were presented and tested by the developer, his colleagues, and coworkers but also by some members of the target user groups to get feedback. For publishing (step five), descriptive meta tags (description, keywords) that can help users on the internet find the game have to be added. Code for web platforms such as Google Analytics can also be implemented in the publishing process to follow user interaction with website games. At the end of the project, when implementing games into Mrežnik, a separate page for all games where the games are categorized will be created so the users can easily browse through them (on previously mentioned link rjecnik.hr/igre). Games have categories and subcategories on dictionary websites based on their content (these categories were decided on in step one). Games will also have tags to connect games of the same type (e.g. memory, quiz, etc.). A unique feature of Mrežnik, which was not found on any dictionary or encyclopedia websites, is that links to some games will be implemented into the entry structure by external links, i.e. certain dictionary entries will have links to games that cover the content of that entry. After the games are published, the next step is to promote them (step six). Promotion can easily and cheaply be done through social media where players can further share the game and even give direct feedback to the developer. In addition to social media, it is also good to have articles written about the games on website portals that are visited by potential players and to promote the games on certain TV shows, in newspapers, and magazines. Another way to promote educational games and also get instant feedback is to have workshops for learners or teachers. All those promotional methods are planned and have already been partly conducted for Mrežnik. The last step (step seven) focuses on monitoring user satisfaction. This can also be done through social media since the users can freely give their opinion. Google Analytics can also be used to check which games or parts of games are being played most often. Questionnaires can also be given to players. Player feedback can be used for creating new games or updating the existing ones and adding new content (which can lead back to step four). Therefore, by explaining and giving examples for all seven steps and creating games based on the content of the born-digital dictionary Mrežnik, we can conclude that it is possible to create a conceptual framework for the gamification of a Croatian born-digital dictionary. The steps of this conceptual framework can also be used for gamifying other lexicographic works. This conceptual framework can also help in developing the methodology for gamifying language content in the digital environment and language games created for Mrežnik can help players in learning Croatian language and dictionary content. The thesis opens further questions that require more research in the future such as: a) how to modify the conceptual framework for gamifying encyclopedias (and possibly archive, museum, and library content); b) how would results of using gamified dictionary content for learning purposes (possibly not online) with a larger sample and different groups (this includes not only doing research on non-native speakers) fair compared to current research results; c) how to implement crowdsourcing for further gamification of lexicographic works; d) with the development of the game industry, will there be a need to identify more educational game types and usable gamification elements in the future. |
