No description available.

<p dir="ltr">Since the Electronic Software Association (ESA) began reporting data for the video game industry in 2002, women have represented nearly half of the game playing population. However, despite this stable statistic, the industry’s ideal “Gamer” is consistently depicted as a young, white, cisgender, heterosexual, able-bodied male, and the games industry frequently targets this idealized identity through advertising and game design. This has resulted in a culture that is notably toxic towards women and marginalized players, built on an assumption of meritocracy within games—or the expectation that every player begins each game with the same advantages, disadvantages, and skills as every other player. While the construction of gamer identity has received extensive scholarly attention, gaming peripherals—such as video game controllers—are either minimized or left entirely out of the conversation. This dissertation, informed by feminist methodologies in technical communication and game studies, uses a mixed-methods approach involving archival research, visual analysis, surveys, and interviews to understand the intersections of video game controllers and gamer identity. Using Microsoft’s Xbox as a case study, the findings demonstrate how a dominant narrative has controlled controller design decisions through iterative processes. This has resulted in controllers that are more uncomfortable, more unusable, and more frustrating for and viewed more negatively by women and marginalized players. For each controller iteration, women and marginalized participants rated controllers significantly lower. Though the total improvement score (TIS) from first iteration to current iteration were similar between women and marginalized participants and cismale participants, the lower starting point for women and marginalized participants resulted in a lower ending point. Design decisions across controller iterations privilege cismale experiences, reifying gamer identity through controller design and resulting in not just an ideal gamer identity, but an ideal gamer body. </p>

This thesis explores the potential of Large Language Models (LLMs) to assist in the technical documentation process at Saab. With the increasing complexity and regulatory demands on such documentation, the objective is to investigate advanced natural language processing techniques as a means of streamlining the creation of technical documentation. Although many standards exist, this thesis particularly focuses on the standard ASD-STE100, Simplified Technical English abbrv. STE, a controlled language for technical documentation. STE's primary aim is to ensure that technical documents are understandable to individuals regardless of their native language or English proficiency.  The study focuses on the implementation of Direct Preference Optimization (DPO) and Supervised Instruction Fine-Tuning (SIFT) to refine the capabilities of LLMs in producing clear and concise outputs that comply with STE. Through a series of experiments, we investigate the effectiveness of LLMs in interpreting and simplifying technical language, with a particular emphasis on adherence to STE standards. The study utilizes a dataset comprised of target data paired with synthetic source data generated by a LLM. We apply various model training strategies, including zero-shot performance, supervised instruction fine-tuning, and direct preference optimization. We evaluate the various models' output using established quantitative metrics for text simplification and substitute human evaluators with company internal software for evaluating adherence to company standards and STE. Our findings suggest that while LLMs can significantly contribute to the technical writing process, the choice of training methods and the quality of data play crucial roles in the model's performance. This study shows how LLMs can improve productivity and reduce manual work. It also looks at the problems and suggests ways to make technical documentation automation better in the future.

This study explores Cummins' interdependence hypothesis in the South African context. The design is experimental, involving Zulu primary language writing instruction to explore whether skills taught in Zulu composition classes transfer into English expository writing. The intervention and control groups were drawn from two ex-Model C high schools and the focus was on measuring use of coherence and cohesion in English essays. Quantitative findings showed, although the intervention group's writing skills did not improve significantly, they did not decline. By contrast, the control group's writing skills declined significantly. A more qualitative investigation of the corpus supports the statistical findings. However, because of the limitations of this study, more research is required into Cummins' hypothesis, bilingual programmes and teaching academic writing skills in African languages. It is hoped that this research design will benefit future researchers investigate the current debate about the efficacy of bilingual and multilingual approaches to education. / Linguistics / M. A. (Applied Linguistics)

In the midst of a period of economic transition, community organizers across Appalachia are working towards a just future that privileges community growth over corporate gain. A recent turn towards social justice concerns in Professional and Technical Communication suggests that efforts of community organizers might be of interest to scholars focused on addressing wicked problems in disenfranchised communities. This dissertation draws from results of a participatory photovoice study in which 11 community organizers took photos, wrote narratives, and responded in focus groups, and site visits to several communities. These methods call for deep engagement with community knowledges, producing rich visual and textual portraits of life in Appalachia that challenge stereotypical renderings of the region and its residents. After providing a heuristic for uncovering and re-valuing community knowledges, this dissertation looks at how place, technology, and community factor into the experiences of community organizers. Results from gathered qualitative data suggest that community members are experts on their own experiences, as participants revealed understandings of complex problems that call into question standard development practices lauded by technical experts. Second, participants demonstrated a capacity for embracing the very elements of their communities that had been used to marginalize them, pointing to the power of unexpected and creative tactics. Lastly, their reflections revealed the need for more attention to be placed upon community organizing in rural contexts and what kinds of community knowledges exist beyond expected parameters. By documenting their experiences organizing around public problems, participants confronted monolithic representations of their region, articulated their own nuanced accounts of life in rural areas, and crafted strategies for community-focused development that privileges people. Ultimately this project argues that by inviting community knowledges into the academic sphere, we might craft more effective coalitions to tackle complex public problems.

Upon IRB approval, a corpus of 1,192 texts consisting of three assignments written by a total of 1,736 first year engineering students was compiled, and 117 pedagogical materials were collected. Using an iterative quantitative-qualitative approach to written discourse analysis, instances of formulaic language (4- and 6-word sequences) were identified in the corpus; formulaic language was then coded for the rhetorical functions expected in problem statements as qualitatively identified in the pedagogical materials. Additionally, three discourse-based interviews were conducted with First-year Engineering Faculty. Interview data was coded for themes of effective communication and used to triangulate the findings from the corpus analysis.

This study explores Cummins' interdependence hypothesis in the South African context. The design is experimental, involving Zulu primary language writing instruction to explore whether skills taught in Zulu composition classes transfer into English expository writing. The intervention and control groups were drawn from two ex-Model C high schools and the focus was on measuring use of coherence and cohesion in English essays. Quantitative findings showed, although the intervention group's writing skills did not improve significantly, they did not decline. By contrast, the control group's writing skills declined significantly. A more qualitative investigation of the corpus supports the statistical findings. However, because of the limitations of this study, more research is required into Cummins' hypothesis, bilingual programmes and teaching academic writing skills in African languages. It is hoped that this research design will benefit future researchers investigate the current debate about the efficacy of bilingual and multilingual approaches to education. / Linguistics and Modern Languages / M. A. (Applied Linguistics)

L’exprérience d’enseignement et l’examen des mémoires en français des étudiants en Agroaliemtaire ont montré que nos apprenants produisaient souvent des phrases longues et peu compréhensibles à cause des erreurs morphosyntaxiques. En conséquence, les mémoires sont peu lisibles sur le plan linguistique. <p>De plus, nous remarquons des traces de la langue vietnamienne dans la production en français des étudiants. Or, les apprenants sont obligés de consulter les articles scientifiques en vietnamien de leurs enseignants lors de la préparation du mémoire. De plus, l’étude des articles montre que les auteurs formulent aussi des phrases très longues de plusieurs informations.<p>Ainsi, toutes ces constatations nous orientent vers une analyse contrastive des phrases longues en vietnamien des scientifiques avec celles trouvées dans les mémoires en français des étudiants. <p>Selon notre revue de littérature des recherches précédentes, des auteurs prédécesseurs mesurent la lisibilité d’un texte en se basant statistiquement sur la familiarité du vocabulaire, la longueur des mots, la longueur des phrases, ou la longueur des sous-phrases. <p>Toutefois, la mesure par le comptage du nombre de mots par phrase des auteurs semble inappropriée à notre travail par la différence des objectifs. <p>Nous avons donc essayé de trouver une unité de mesure de la longueur des phrases pertinente à notre propre corpus :« informations enchâssées ».<p>Selon les auteurs prédécesseurs, une phrase sera vue comme longue si elle contient plus de trois sous-phrases. Quant à nous, les phrases seront jugées longues si elles dépassent trois informations enchâssées. <p>Après la collecte des phrases longues, nous avons utilisé l’approche qualitative pour les analyser. Après l’analyse du corpus, nous avons obtenu des résultats suivants :la production des phrases longues ainsi que la présence des erreurs morphosyntaxiques dues à l’interférence du vietnamien constituent des caractéristiques typiques des mémoires des étudiants francophones à Can Tho. Ce sont ces traits représentatifs qui ont compromis la lisibilité des phrases de nos apprenants.<p>Face aux difficultés de nos apprenants, nous essayons de trouver quelques esquisses didactiques adéquates à notre propre public.<p> / Doctorat en Langues et lettres / info:eu-repo/semantics/nonPublished