While block-based programming has successfully eliminated critical syntactic barriers to programming, it remains unclear how effectively it aids in overcoming semantic, logical, and pragmatic programming challenges that hinder computational thinking. These challenges are likely to far outweigh the syntactic ones. With the goal of creating a highly accessible programming tool for young students using mobile devices, we explored the concept of pragmatic prebugging to begin addressing these challenges. By pragmatic prebugging, we refer to proactive debugging tools designed to prevent logical errors. This article introduces RULER.game as a Computational Thinking Tool with built-in pragmatic prebugging, enabling novice programmers to create games through a paradigm we call proxy-based programming. A small study exploring error rates found statistically significant performance improvements of proxy-based programming compared to block-based programming.

Block-based programming languages effectively address syntactic difficulties allowing users to more easily create code. Syntactic code support is no doubt a crucial step in enabling the next generation of programmers. However, in what other ways do these tools support the computational thinking (CT) process? For example, how is CT supported in both the semantic and pragmatic levels? Most users do not just want to write code?they want to create interesting artifacts such as video games, simulations, stories and robots. No doubt that users must code to build these artifacts, but just as important, though often overlooked, is that users must also have an understanding of tool affordances to effectively build these artifacts. The position presented in this paper is that the comprehension of affordances is essential to each stage of the computational thinking process. To that end, we introduce the Hourglass Challenge as a way for students to experience affordances through a programming challenge based on mis-aligned tools and problems. The results from a study, with 54 preservice elementary school teachers, indicate that the affordances of tools can have profound consequences on computational thinking. For instance, the affordances various tools provide can be the difference between creating a simulation versus an animation. But just as hammers are not intrinsically better tools than screwdrivers, the contribution of this position paper is not to compare tools, but to promote efforts that enable learners to develop a better understanding of computational thinking tool affordances as essential mediators between problems and tools.

Recently, Computer Science education in Switzerland reached a critical milestone. In the past two years the School of Education of FHNW, one of Switzerland's largest schools of education, introduced a mandatory Computer Science education module, comprised of a two-course sequence for its students: pre-service primary-level schoolteachers. Due to its design and audience, this mandatory module introduces a variety of unique challenges. The module is taken by a majority of female students (75%), many with no prior experience and/or interest in programming and Computer Science. The module starts with a course focused on learning Computational Thinking, referred to as FW, followed by a course focused on teaching Computational Thinking, referred to as FD. Since 2017, 1000 pre-service primary-level teachers have been trained in Computer Science education through this module. This curricula initiative computes the effect sizes from teaching two complete iterations of the "FW then FD" design. It is based on attitude assessment data collected at four points: Pre-FW, Post-FW, Pre-FD, Post-FD, during the first two years of this mandatory module. Our research indicates that, by course's end, pre-service teachers are mostly confident in their ability to program and think computationally.

A promising approach to make K-12 Computer Science education more systemic could arise from a strategy focusing mostly on pre-service teachers educated through mandatory courses instead of self-selected in-service teachers. When employing mandatory courses, schools of education can reach all future teachers, but what are potential consequences resulting from this demographic shift? Pre-service teachers may not expect to acquire programming skills and may not be convinced of the relevance of Computer Science. In 2017, one of the first mandatory Computer Science education courses for pre-service K-12 teachers was introduced at the School of Education of northwestern Switzerland (PH FHNW). The mandatory nature of the course was possible because of the introduction of Computer Science as a subject in a new national curriculum. The course, based on Scalable Game Design, was taken by over 600, mostly female (75%), pre-service elementary school teachers. This paper explores the characteristics of this new audience and investigates the consequences of mandatory pre-service teacher Computer Science education. While our research shows that the course was successful, with regards to improving the students' skills, it reveals significant gender effects concerning attitudes towards Computer Science and self-efficacy.

The creation of computer science tutorials is becoming critically important as hundreds of millions of students each year get their first CS experience through self-directed online activities. Creating a "cliffhanger" activity, with high engagement during and motivation to continue learning post activity, is a balancing act. If tutorials provide too much detailed information, users may be able to follow instructions but can feel overwhelmed or bored. On the other hand, tutorials that do not sufficiently explain crucial steps risk frustrating users who might drop out of the activity. Zones of Proximal Flow (ZPF) tutorials are simple to create and provide a navigation structure of differentiated instruction allowing users to choose appropriate detail based on their self-assessed state of flow, from bored to anxious. Using Retention of Flow analysis, two Hour of Code game design tutorials were analyzed: a sophisticated online tutorial for the creation of Frogger, and a simple ZPF tutorial for the creation of Pac-Man. One hope was that the simple ZPF Pac-Man tutorial would not do much worse than the sophisticated Frogger tutorial, but surprisingly the ZPF Pac-Man tutorial significantly outperformed the Frogger tutorial in terms of student retention. The Pac-Man tutorial also displayed a high student motivation to continue programming past the end of differentiated instruction.

Broadening participation in computer science necessitates balancing motivational and educational concerns. Without fully understanding potential trade-offs, Hour of Code-like tutorials may actually backfire by initially attracting students to participate, but gradually reinforcing the notion that programming is hard and boring. Previously, we analyzed and compared two Hour of Code tutorials: a tutorial that walks students through the creation and programming of a 3D-Frogger game, including the drawing of their own 3D characters, to a programming puzzle where students solve discrete coding challenges. Using an analysis based on retention, the comparison indicated higher levels of perseverance in the game creation activity. However, does the ability to draw characters really motivate students to program? Conflicting theories of positive and negative effects of drawing onto perseverance include that drawing might increase levels of participant ownership or that drawing may just be wasting time better spent programming, especially in the time constrained Hour of Code context. To gain insight, this study uses draw times of over 8,000 projects from a game creation Hour of Code activity in 2016. Initial results indicate that higher average draw time per character corresponds to increased program lengths and students with the highest average draw times per character continued to program beyond the end of the activity

Computational Thinking is argued to be an essential skill for the workforce of the 21st century. As a skill, Computational Thinking should be taught in all schools, employing computational ideas integrated into other disciplines. Up until now, questions about how Computational Thinking can be effectively taught have been underexplored preventing efforts to cross the large gap between early adopters and the early majority, conceptualized as the Computer Science Education chasm. A promising strategy to cross the chasm is underway in Switzerland. Switzerland recently introduced a national curriculum, called Lehrplan 21, mandating Computer Science Education. This mandate requires the Computer Science education of elementary and middle school students. In 2017, the School of Education of Northwestern Switzerland (PH FHNW), introduced a mandatory pre-service teacher Computer Science Education course, to satisfy this mandate. All the PH FHNW students who study to become elementary school teachers must pass this two-semester course. The first part of this course was taught for the first time in fall of 2017. This paper presents the philosophy of this course and an initial analysis of both qualitative data capturing the students’ perceptions of Computational Thinking and quantitative data describing shifts in students’ skills and attitudes as effect sizes. The data suggest that it is possible to teach a basic understanding of programming to non-self-selected pre-service elementary school teachers.

Currently Switzerland is going through a major reform in its education system. One of its most ambitious and important goals is the inclusion of Computer Science Education already on the primary school level, an important measure in achieving the establishment of an information society. Such a reform raises questions about the appropriate types of approaches to be developed and employed for an effective implementation of Computer Science Education concepts in Swiss primary schools. To this end, the project "Scalable Game Design Solothurn" was developed and evaluated. This project both trained teachers and exposed students to Computational Thinking concepts through the two Computational Thinking Tools AgentSheets and AgentCubes online. Results show that teaching Computational Thinking through Scalable Game Design is not only feasible on the primary school level but also enjoyable, with AgentSheets and AgentCubes online proving to be sustainable and effective tools for the implementation of Computer Science Education on this school level. Further analysis of the data enables us to make recommendations regarding optimal ways of implementation for the Swiss reality and point towards new research directions.

Online CS Ed Week and Hour of Code activities attempt to motivate hundreds of millions of student participants across the world in computer science each year. A key goal of these endeavors is long-term student engagement. However, if the activity experience is bad, it could have effects adverse to the stated goal. Thus, it is imperative upon designers to actively improve the online activity ensuring the maximum numbers of students are retained throughout the exercise. We present a simple proof of concept method outlining a means for Computer Science Education Week and Hour of Code online activities to identify and improve hazardous points wherein students tend to drop out. This is achieved by finding so called flow stoppers in activity retention that diverge from an ideal theoretical Markov chain model, and scaffolding the activity at that point to better support participants. Initial data presented indicates that even minor changes can have a significant effect on keeping a greater number of students engaged.

Computer Science Education Week activities, featuring online? programming tools embedded with tutorials, report large participation numbers. However, to truly broaden participation, activities need to be made accessible in international contexts. In 2014, Tecnológico de Monterrey and Instituto de Innovación y Transferencia de Tecnología de Nuevo León, modified the Scalable Game Design CS Ed Week activity to include a Mexican feasibility pilot study. The goal of the pilot was to broaden participation in Computer Science in Mexico by creating interest and demand in further activities, including launching of 2015 Mexico CS Ed Week. This paper reviews the initial results of this 2014 pilot, including the discussion of the unique challenges faced in this context, and examines efforts to make this activity more accessible and successful. In addition to pilot data highlighting future activity improvements, initial retention results show that despite challenges, Mexican students were able to effectively use the modified activity to create games on par with U.S. students.

High profile computer science education events such as the Hour of Code can reach millions of students but without proper evaluation it is not clear what motivational and educational consequences the participation has. If, for instance, participants' levels of motivation towards the end of an hour long activity are significantly fading, then their perception of programming to be "hard and boring" may actually get reinforced. By simply measuring how far participants progressed with their projects we have been able to collect retention data from thousands of participants in a way that allows us to interpret these data in terms of not only cognitive but also technical and practical activity challenges. Inflection points overlaying a negative exponential retention distribution serve as indicators of these challenges with potential impact on Flow. Retention of Flow is an evaluation approach to analyze computer science education activities, including interactive tutorials and online programming environments, with respect to cognitive as well as affective challenges.

Computational Thinking is an essential skill for all students in the 21stCentury. A fundamental question is how can we create computer affordances to empower novice teachers and students, in a variety of STEM and art disciplines, to think computationally while avoiding difficult overhead emerging from traditional coding? Over the last 20 years we have iteratively developed tools that aim to support computational thinking. As these tools evolved a philosophy emerged to support Computational Thinking by joining human abilities with computer affordances. Chief among these findings is that supporting Computational Thinking is much more than making coding accessible. Computational Thinking Tools aim to minimize coding overhead by supporting users through three fundamental stages of the Computational Thinking development cycle: problem formulation, solution expression, and solution execution/evaluation.

Computational thinking (CT) involves a broadly applicable and complex set of processes that are often explained by way of the knowledge, attitudes, and general practices that they entail. However, to become facile with CT, learners require instruction that is grounded in concrete, relevant experiences. This paper examines teacher practices that are intended to promote CT skill acquisition through instruction that takes place in two framing contexts. The phenomenological context, which is based on observable patterns of object interaction that recur in games and simulations, is particularly valuable for developing the capacity to think abstractly. Abstraction is the key to recognizing analogous conditions, an ability that is the basis for transferring learning to new situations. The disciplinary context describes areas of application within and across subject areas, including computer science, that can foster proficiency with data representation, problem decomposition, and other CT skills. Using the Scalable Game Design curriculum as a lens to examine classroom practices, we find that teachers both plan and enact CT instruction in these contexts.

Teachers are increasingly integrating game design and simulation creation projects as part of their classroom curricula. These projects have many benefits including motivating students in STEM activities and exposing students to computational thinking - a key part of upcoming science standards. However, barriers still exist to project-based computer science in a lab environment. One major issue is that, as students are creating their projects, it is extremely difficult for teachers to know how every student is progressing through a given activity and how to keep every student engaged. This paper introduces a Cyberlearning system entitled REACT (Real-Time Evaluation and Assessment of Computational Thinking) that is an initial step towards giving teachers quickly discernible real-time data corresponding to each student project. REACT provides teachers with a sortable dashboard, consisting of data from each student, that shows the characters students created and used to populate their game or simulation world as well as the semantic meaning behind what students have programmed. A feasibility test with four middle school classrooms shows that REACT helps teachers formatively assess students and provide targeted instruction to struggling individuals. Furthermore, teachers showed excitement at the summative and student self-assessment capabilities of REACT, and every teacher independently stated they would use the REACT system in subsequent end-user programming units.

Building a classroom environment based on student-led, meaningful collaboration is an ideal being promoted by educators and administrators seeking to equip students with 21st century workforce skills. More and more often, teachers are challenged to design lessons that utilize students' innate desire to communicate with one another rather than more traditional direct instruction. With a heavy focus on student-driven instruction and a reduced role for direct instruction, Carson Middle School's Game Design I course is able to utilize collaboration as a means for allowing students to not only learn but master and retain Computational Thinking Patterns and apply them in formal summative assessments. Through the collection of these data points throughout the course of a semester of learning, an educator can fully appreciate the learning of Computational Thinking skills that occurs as students build a sequence of games and simulations.

An educated citizenry that participates in and contributes to science technology engineering and mathematics innovation in the 21st century will require broad literacy and skills in computer science (CS). School systems will need to give increased attention to opportunities for students to engage in computational thinking and ways to promote a deeper understanding of how technologies and software are used as design tools. However, K-12 students in the United States are facing a broken pipeline for CS education. In response to this problem, we have developed the Scalable Game Design curriculum based on a strategy to integrate CS education into the regular school curriculum. This strategy includes opportunities for students to design and program games and science technology engineering and mathematics simulations. An approach called Computational Thinking Pattern Analysis has been developed to measure and correlate computational thinking skills relevant to game design and simulations. Results from a study with more than 10,000 students demonstrate rapid adoption of this curriculum by teachers from multiple disciplines, high student motivation, high levels of participation by women, and interest regardless of demographic background.

End-user game design affords teachers a unique opportunity to integrate computational thinking concepts into their classrooms. However, it is not always apparent in game and simulation projects what computational thinking-related skills students have acquired. Computational Thinking Pattern Analysis (CTPA) enables teachers to visualize which of nine specific skills students have mastered in game design that can then be used to create simulations. CTPA has the potential to automatically recognize and calculate student computational thinking skills, as well as to map students' computational thinking skill progression, as they proceed through the curriculum. The current research furthers knowledge of CTPA by exploring its validity based on how its performance correlates to human grading of student games. Initial data from this validation study indicates that CTPA correlates well with human grading and that it can even be used to predict students' future achievement levels given their current skill progression, making CTPA a potentially invaluable computational thinking evaluation tool for teachers.

Future school science standards, such as the Next Generation Science Standards (NGSS), emphasize the integration of simulation and modeling activities in the classroom environment. The extremes of these activities have two vastly different implementations. On one hand, a teacher can have students experiment on a pre-made simulation associated with the material. On the other hand, students can use, for example, an end-user programming tool to create the simulation from scratch. This allows students to not only experiment on, but also, to model the real world phenomenon being studied- a key component of computational thinking. However, the greater amount of time necessary for student authoring of simulations can make such an approach infeasible in the classroom environment. This paper presents a spectrum of strategies for integrating simulations into class- rooms emphasizing our research at the Scalable Game Design Lab, University of Colorado Boulder as well as research from other entities. Starting at consuming simulations and adding more user interaction and authoring elements begins to provide a gentle slope from consumption towards simulation creation. Results indicate that many of these strategies are quite effective.

This paper suggests a Cyberlearning tool based on a highly innovative assessment methodology that helps teachers with computer science education. Currently, there is a strong push to integrate aspects of programming and coding into the classroom environment. However, few if any tools exist that enable real-time formative assessment of in-class programming tasks. The proposed REACT (Real Time Evaluation and Assessment of Computational Thinking) system is a first step toward allowing teachers to see which high-level concepts students have mastered and which ones they are struggling with as students code in real time. REACT supports and facilitates the teaching of 21st century computing skills such as computational thinking [1] in the classroom environment.

This paper presents a novel pedagogical framework, entitled the Zones of Proximal Flow, which integrates Vygotsky's Zone of Proximal Development theory with Csikszentmihalyi's ideas about Flow. Flow focuses on the individual-- an individual is in Flow when challenges are balanced with skills. The Zone of Proximal Development, on the other hand, brings in a social learning aspect focusing on a student's ability to learn concepts with external support. From our research experiences bringing game and simulation design into middle school classrooms, we attempt to provide students with appropriate challenges using a project-first based approach that aims to keep students in Flow. The project-first approach employs inquiry based scaffolding to guide students, with appropriate support by their teachers, through Vygotsky's Zone of Proximal Development, back in to Csikszentmihalyi's state of Flow for an ideal learning experience. We call this space the Zones of Proximal Flow. Data indicate that the Zones of Proximal Flow approach works, keeping classrooms engaged in the act of game design and enabling students to advance to more complex program creations.

Computational thinking aims to outline fundamental skills from computer science that everyone should learn. These skills include problem formulation, logically organizing data, automating solutions through algorithmic thinking, and representing data through abstraction. One aim of the NSF is to integrate these and other computational thinking concepts into the classroom. This paper introduces a tool called the Simulation Creation Toolkit wherein users apply high-level agent interactions called Computational Thinking Patterns (CTPs) to create simulations. Programming at the Computational Thinking Pattern level allows users to directly create agent interactions in a simulation by employing generic icons acting out a scientific phenomenon they are trying to represent. The Simulation Creation Toolkit aims to preserve the computational thinking benefits of simulation creation while enabling higher-level implementation of agent behaviors. Initial study data collected from sixth grade students with no prior programming experience indicates that students can work the mechanics necessary to create simulations in the classroom environment using the Simulation Creation Toolkit.

Conventionally creativity is often conceived as an aptitude to be discovered in an individual that cannot be mathematically measured. But the concept of creative thinking as a divergence from a standard "norm" is used in creativity research for the purpose of assessing creativity and is also linked to non-traditional or creative processes that lead to unique and divergent artifacts [1,2]. Using Computational Thinking Pattern Analysis (CTPA)[3], the divergence between implemented computational thinking patterns in a student-created game, and that game's tutorial "norm" is calculated as an indicator of creativity. Through a case study of one teacher using three unique learning conditions, CTPA's computed divergence is explored as a valid measurement of creativity in these student games.

A strategy exposing middle school students to computer science through game design appears to be a promising means to mitigate the computer science pipeline challenge. Particularly when short game design activities are integrated into already existing middle school courses, research suggests that game design is effective in broadening participation and motivating large numbers of students, along with large percentages of women and minorities. A study with over 10,000 students is exploring the sustainability of this approach and finding positive responses to inquiries such as these: Do teachers continue to use game design? Can they advance beyond extrinsic rewards such as research stipends? After building one game, do students advance, building more games or even STEM simulations?

Educational game design environments have long been employed as an engaging medium for teaching computer programming and software engineering concepts such as objects, agents, instances, behavior, and interaction. However, in contrast to the collaborative nature of programming, in particular among novice programmers, existing game design environments are solitary desktop-based applications with no or very limited support for sharing and collaboration. We have taken advantage of Web 2.0 technologies and its participative culture to develop a collaborative platform around game design activities that encourages social learning and integrates into the students' online social presence and lifestyle. We have employed open-Web technologies to build a game design environment using only HTML5 and JavaScript that lets users practice the game design and programming activities right inside their browser. Game design and game playing run on the client while an associated website hosts the created games and supports collaboration. This paper reports on the design, implementation, and evaluation of the game design environment. The preliminary evaluation of Agent Web's usability and performance demonstrates that not only does Agent Web meet its design goals, but also that the current Web is a hospitable environment for end-user game programming both on desktop and mobile devices.

A fundamental challenge to computer science education is the difficulty of broadening participation of women and underserved communities. The idea of game design and game programming as an activity to introduce children at an early age to computational thinking in a motivational way is quickly gaining momentum. A pedagogical approach called Project First has allowed the Scalable Game Design project to reach a large group of middle schools students including a large percentage of female (45%) and underrepresented (48%) students. With over 4000 students in inner city, remote rural, and Native American communities Scalable Game Design has investigated the impact on students' interest level of pedagogical approaches employed by teachers such as mediation and scaffolding. Findings suggest strong gender effects based on classroom scaffolding approaches. For instance, girls are substantially less likely to be motivated through scaffolding based on direct instruction. Conversely, guided discovery scaffolding approaches are highly motivating to the point where they can even overcome other negative predictors such as small girls to boys class participation ratios. This paper introduces the project, discusses different scaffolding approaches and presents data connecting gender specific motivational levels with scaffolding approaches.

Divergent thinking has been linked to creative processes leading to innovative artifacts. Measuring creative divergence can be difficult. Across the USA, the Scalable Game Design (SGD) Project includes thousands of student participants building their own games through learning computational thinking (CT). To evaluate these games, a technique, the Computational Thinking Pattern Analysis (CTPA) [1], was developed, refined and used successfully. Under three different learning conditions, divergence was computed through CTPA, and then analyzed and explored as an indication of creativity.

Creativity is an important aspect of industry and education. The lack of creativity in current students has become a concern for educators. Through the process of implementing the Scalable Game Design project to teach computer science through game authoring, fostered/increased creativity occurred in public middle schools. Despite some structural limitations of the US educational system, creativity among the participating students was recognized. This paper describes a unique solution to fostering creativity while teaching game design in the limiting public school environment.

End-user game design tools are effective in motivating and exposing students with no prior programming experience to computer science. However, while there is good evidence that these environments are effective motivators, the question remains what do students actually learn? For our purposes, using AgentSheets, we would like to know if students can apply the knowledge obtained from programming games to creating science simulations. Specifically, we want to better understand if students are able to recognize Computational Thinking Patterns (CTP) from their game programming experience. Computational Thinking Patterns are abstract programming patterns that enable agent interactions not only in games but also in science simulations. Students and teachers who participated in a game design summer institute were administered a Computational Thinking Pattern Quiz (CTP Quiz). This quiz tested the participants' ability to recognize and understand patterns in a context removed from game programming. We found that participants, for the most part, were able to understand and recognize the patterns in a variety of contexts

How can learning be computed? Curriculum, using visual language as the motivational context with embedded computer science content was utilized in one college computer science class and two middle school technology classes. From the data collected in these three classes over the course of a semester, associated learning progressions were computed from several computational thinking patterns. By comparing the results (learning progressions), some obvious and some not so obvious indications emerged. We believe that the more obvious indications give credence to the less obvious, and hence the measurement tool. This comparison between middle school students and college students demonstrated that middle school students' learning progressions are slower than college level students (obvious), but middle school students' learning skill scores reached/surpassed the entry level scores of students in the college class after designing only two or more games. Consequently, these results tell us that this measurement tool although not validated yet, could measure accumulated learning in certain contexts. Although this is just the first "outing" of this tool these results strongly indicate that the tool provides an accurate measure of the concept, in this case, learning accumulation or transfer. The Scalable Game Design (SGD) project was funded by the National Science Foundation (NSF) to improve computer science interest in middle schools. The goal of SGD is to use visual language programming software (Agentsheets) to introduce a more positive image of computer science through a game design course.

Summary form only given. Above showsthat the next generation of mobile applications for MOPs need to be part of a much large mobile information system. Such system should be capable of receiving information from a large number of users, intelligently aggregating the information to useful, meaningful, real-time information and making them available to users. Users should be able to make queries and quickly find information that would enhance their day to day activities. It should also have the functionalities to detect different and unusual situations, provide alerts to users in unusual circumstances and emergencies so that users can respond accordingly. These next generation mobile applications for MOPs will open up a very large market currently estimated to be around 3 billion people [13]. These applications will empower them to be information users as well as information producers. Such applications would result in up lifting of economic conditions of the MOPs and transformation of social processes.

Accelerated by the Do-It-Yourself mindset of the Web 2.0 culture, end-user programming-programming by end users with limited or even no formal programming background-is growing rapidly. Especially in educational settings, children are exposed to computational thinking by making games, building scientific simulations and creating stories. Early educational programming languages such as Logo have made programming substantially more accessible to end-users. More recent approaches include visual programming with a drag-and-drop style of programming that makes it nearly impossible to compose syntactically incorrect programs. However, as the syntactic challenges of end-user programming are gradually fading into the past, the new frontier of semantic programming support emerges. This demonstration introduces Conversational Programming, a system to make programming more conversational. A conversational programming agent runs programs one step into the future in order to help end-users visualize discrepancies between the programs they intended to write and their actual programming results.

Accelerated by the Do-It-Yourself mindset of the Web 2.0 culture, end-user programming, which is programming by end users with limited, if any, formal programming background, is growing rapidly. Especially in educational settings, children are exposed to computational thinking by making games, building scientific simulations and creating stories. Early educational programming languages such as Logo have made programming substantially more accessible to end users. More recent approaches include visual programming with drag-and-drop style of programming making it nearly impossible to compose syntactically incorrect programs. However, as the syntactic challenges of end-user programming are gradually fading into the past, the new frontier of semantic programming support emerges. This demonstration introduces Future Trace, a system to make programming more conversational. A conversational programming agent runs programs one step into the future in order to visualize discrepancies between the programs users intended to write and the actual programs.

A variety of approaches exist to teach computer science concepts to students from K-12 to graduate school. One such approach involves using the mass appeal of game design and creation to introduce students to programming and computational thinking. Specifically, Scalable Game Design enables students with varying levels of expertise to learn important concepts relative to their experience. This paper presents our observations using Scalable Game Design over multiple years to teach middle school students, college level students, graduate students, and even middle school teachers fundamental to complex computer science and education concepts. Results indicate that Scalable Game Design appeals broadly to students, regardless of background, and is a powerful teaching tool in getting students of all ages exposed and interested in computer science. Furthermore, it is observed that many student projects exhibit transfer enabling their games to explain complex ideas, from all disciplines, to the general public.

Cyberlearning infrastructures are increasingly being integrated into physical classrooms and are often used by online classes as an outright replacement for the physical classroom. In Spring 2009, The Educational Game Design Class, taught at the University of Colorado Boulder, employed a cyberlearning infrastructure enabling students to run and even download classmate assignment submissions before the homework deadline. This cyberlearning infrastructure, called the Scalable Game Design Arcade, also allowed students to give feedback on other students' assignments. Analysis of data from the Scalable Game Design Arcade indicates that students used the online infrastructure to play and appraise fellow students' games. Interestingly however, data suggests that most students preferred to give feedback verbally in-person during class; data also indicates that in-class feedback was the most effective in terms of getting students to improve and resubmit their assignments.

Game design appears to be a promising approach to interest K-12 students in Computer Science. Unfortunately, balancing motivational and educational concerns is truly challenging. Over a number of years, we have explored how to achieve a functional balance by creating a curriculum that combines increasingly complex game designs, computational thinking patterns and authoring tools. Scalable Game Design is a research project exploring new strategies of how to scale up from after school and summer programs into required curriculum of public schools through game design approaches. The project includes inner city schools, remote rural areas and Native American communities. A requirement checklist of computational thinking tools regarding curriculum, teacher training, standards and authoring tools has been developed and is being tested with thousands of students.

Visual programming languages can be used to make computer science more accessible to a broad range of students. The evaluative focus of current research in the area of visual languages for educational purposes primarily aims to better understand motivational benefits as compared to traditional programming languages. Often these visual languages claim to teach students computational thinking concepts; however, although the evaluations show that students may exhibit more enthusiasm, it is not always clear what computational thinking concepts students have actually learned. In this paper we attempt to develop a visual semantic evaluation tool for student-created games and simulations that goes towards depicting the computational thinking concepts implemented by the students. Through semantically analyzing a given student's created projects over time, this visual evaluation tool, called the Computational Thinking Pattern (CTP) graph, can possibly indicate the existence of computational thinking transfer from games to science simulations.

Computer visualization and networking have advanced dramatically. 3D hardware acceleration has reached the point where even low-power handheld computers can render and animate complex 3D graphics efficiently. Unfortunately, end-user computing does not yet provide the necessary tools and conceptual frameworks to let end-users access these technologies and build their own networked interactive 2D and 3D applications such as rich visualizations, animations and simulations. The Agent Warp Engine (AWE) is a formula-based shape-warping framework that combines end-user visualization and end-user networking. AWE is a spreadsheet-inspired framework based on Web sharable variables. To build visualizations, users define these variables, relate them through equations and connect them to 2D and 3D shapes. In addition to basic shape control such as rotation, size, and location, AWE enables the creation of rich shape warping visualizations. We motivate the AWE approach with the Mr. Vetro human physiology simulation supporting collaborative learning through networked handheld computers.

Computer visualization has advanced dramatically over the last few years, partially driven by the exploding video game market. 3D hardware acceleration has reached the point where even low-power handheld computers can render and animate complex 3D graphics efficiently. Unfortunately, end-user computing does not yet provide the necessary tools and conceptual frameworks to let end-user developers access these technologies and build their own interactive 2D and 3D applications such as rich visualizations, animations and simulations. In this paper, we demonstrate the Agent Warp Engine (AWE), a formula-based shape-warping framework for end-user visualization.

Game development is quickly gaining popularity in introductory programming courses. Motivational and educational aspects of game development are hard to balance and often sacrifice principled educational goals. We are employing the notion of scalable game design as an approach to broaden participation by shifting the pedagogical focus from specific programming to more general design comprehension. Scalable game design combines the Flow psychological model, the FIT competency framework and the AgentSheets rapid game prototyping environment. The scalable aspect of our approach has allowed us to teach game design in a broad variety of contexts with students ranging from elementary school to CS graduate students, with projects ranging from simple Frogger-like to sophisticated Sims-like games, and with diverse cultures from the USA, Europe and Asia.

Web 2.0 has enabled end users to collaborate through their own developed artifacts, moving on from text (e.g., Wikipedia, Blogs) to images (e.g., Flickr) and movies (e.g., YouTube), changing end-userpsilas role from consumer to producer. But still there is no support for collaboration through interactive end-user developed artifacts, especially for emerging handheld devices, which are the next collaborative platform. Featuring fast always-on networks, Web browsers that are as powerful as their desktop counterparts, and innovative user interfaces, the newest generation of handheld devices can run highly interactive content as Web applications. We have created RistrettoMobile, a Web-compliant framework for running end-user developed applications on handheld devices. The Web-based RistrettoMobileincludes compiler and runtime components to turn end-user applications into Web applications that can run on compatible handheld devices, including the Apple iPhone and Nokia N800. Our paper reports on the technological and cognitive challenges in creating interactive content that runs efficiently and is user accessible on handheld devices.

With the IT crisis reaching alarming levels, it is more important than ever to attract K-12 students to computer science. 3D game development can be an enticing way to achieve that, but building 3D games is far from trivial. Students need to achieve a degree of 3D fluency in modeling, animation and programming to be able to create compelling 3D content. The combination of innovative end-user development tools and standards-based curriculum promoting IT fluency by shifting the pedagogical focus from programming to design, can address motivational aspects without sacrificing principled educational goals. The AgentCubes 3D game-authoring environment raises the ceiling of end-user development without raising the threshold. Our formal user study shows that with Incremental 3D, the gradual approach to transition from 2D to 3D authoring, middle school students can build sophisticated 3D games including 3D models, animations and programming.

Increasingly, public transportation systems are equipped with Global Positioning Systems (GPS) connected to control centers through wireless networks. Controllers use this infrastructure to schedule and optimize operations and avoid organizational problems such as bunching. We have employed this existing infrastructure to compute highly personalized information and deliver it on PDAs and cell phones. In addition to guiding people using public transportation by showing them which bus they should take to reach specific destinations, we track their location to create spatial awareness to a community of users. An application of this technology, called Mobility Agents, has been created and tested for people with cognitive disabilities. About 7% of the U. S. population has a form of cognitive disability. Cognitive disabilities are limitations of the ability to perceive, recognize, understand, interpret, and respond to information. The ability to use public transportation can dramatically increase the independence of this population. The Mobility Agents system provides multimodal prompts to a traveler on handheld devices helping with the recognition of the "right" bus, for instance. At the same time, it communicates to a caregiver the location of the traveler and trip status. This article describes our findings at several levels. At a technical level, it outlines pragmatic issues including display issues, GPS reliability and networking latency arising from using handheld devices in the field. At a cognitive level, we describe the need to customize information to address different degrees and combinations of cognitive disabilities. At a user interface level, we describe the use of different mission status interface approaches ranging from 3D real-time visualizations to SMS and instant messaging-based text interfaces.

Now that we have end-user programming environments capable of empowering kids with no programming background to build games in a matter of hours, a new quest for raising the ceiling of end-user development is emerging. Environments not only focusing on programming, but also including rich media such as 3D, could work as compelling tools for introducing information technology at the K-12 level. The new challenge is raising the ceiling without raising the threshold. Based on our experience with AgentSheets, a tool used worldwide for computational science and game design applications, we created a new authoring tool called AgentCubes. This article discusses the notion of incremental 3D as a design approach for media-rich end-user development with low threshold and high ceiling in education