It is challenging to retain computing students through their first stage of undergraduate education. Attrition is high, with many transferring courses or dropping out. This poster explores preliminary findings from an action research project improving continuation in first-stage undergraduate computing. Five years of data from Falmouth University's Games Academy in the UK suggest improvement in first-stage retention from 66.6% in 2017-18 to 91.2% in 2021-22. Findings support prior work on pair programming, media computation, and peer instruction. However, they also highlight the benefits of collaborative learning facilitated by faculty and informed by learning analytics. Peer reviews and pre-submission clinics, student advisor follow-ups, and retrieval via synoptic assessment also contributed to the improvement.

In this paper, the authors explore the development process for Tregeagle's Kerdroya, an augmented reality app designed to immerse players in the exploration of a real physical labyrinth on Bodmin Moor. The authors discuss the lineage of this game in previous projects from Falmouth University and WAVE, contextualizing Kerdroya within a larger set of themes involving co-creation and educational game design. The authors further unpack Kerdroya's complex design and development process, situating it within the context of debates surrounding collaborative authorship, co-design, and co-creativity. The authors then explore the specific design and development process of Kerdroya with emphasis on the strengths and weaknesses associated with collaboratively developing an experimental app whose aesthetic sensibilities run counter to many of the assumptions of commercial apps regarding usability and polish. The authors conclude that the financial and time constraints of the process allowed them to make a virtue of necessity, finding opportunities for avant-garde creativity amid scarce resources and limited time.

Computing has long been considered to have poor retention rates especially at a CS1 level. This research aims to address this by providing a novel support mechanism developed to teach CS1 students debugging skills and measuring the impact this has upon the retention rates of the participants.

MIST, the Mathematical Image Synthesis Toolkit, is an open-source, Web-based graphics application in which students develop skills in computational thinking and deepen their understanding of mathematical functions through creative play in making images. Using a drag-and-drop interface, students build functions that describe animated abstract images. These functions combine a few basic images (e.g., a horizontal blend, a vertical blend, and a time-based blend) using mathematical operators whose semantics are based on corresponding points in each image (e.g., MIST includes models for adding and multiplying colors). Students develop a variety of skills in computational thinking by using the small set of images and operations to develop complex images, not only images that the system challenges them to make but also images that they find themselves inspired to create and even images that their peers challenge them to build. MIST also incorporates a community of artists/learners. Images created in MIST are available to other MIST users to comment on and remix (provided the designer permits reuse). MIST community members may also create challenges for other community members and can curate "galleries" of images they find particularly engaging, interesting, or even useful as challenges for others. Users can also share their images via Facebook or "tweet" their recent creations. We find that this additional community further encourages students to develop new and interesting images.

Consider the following scenario: A student is browsing Facebook and sees a strangely compelling image on her friend's page. She follows the "how I made this link" to learn more. She finds herself on a page with a gallery of animations that she browses. Then she sees a "Create" button and clicks it. She skims the instructions and decides that this is not a "standard" graphics program - rather than using a timeline and "tools", it has just a few basic images (e.g., a horizontal blend, a vertical blend, some time-blends) and a few operations that you use to build new images from existing images. She plays a bit. And she finds that she can create some strange and interesting images, but not necessarily the ones she wants to make, and certainly not anything like her friend made. So she looks further. She learns that there are "challenges" that help you learn the system (and, as importantly, that help you consider and master different aspects of computational thinking), and tries to figure out how to make images as seemingly simple as a triangle or a circle using the limited selection of basic images and operations. It's not easy, but she finds it fun to try (and sometimes more fun to fail -- failures also create interesting images). You've just read a sample encounter with the Mathematical Image Synthesis Toolkit, or MIST. MIST is an open-source, Web-based graphics application that takes a constructionist approach in which open experimentation in image making helps students develop skills in computational thinking and deepen their understanding of mathematical functions. MIST is available at http://glimmer.grinnell.edu.