PhD Thesis Defense: William Rebelsky, Working Towards Universal Computational Literacy

Content

Speaker:

William Rebelsky

Abstract:

The content of math and science courses needs to change to include technological advances in data science and artificial intelligence. Computational literacy, which directly supports scientific thinking, data analysis, and working with complex systems, should be included. This dissertation divides the skills that fall under computational thinking into four overall categories and describes our work towards improving what students learn and how they engage in problem-solving.

To be computationally literate, students must be mathematically literate. Given both the decrease in mathematical ability in the U.S.A over time and the fact that middle school is a core time to address this mathematical decline, we created and evaluated a process to automatically generate math problems and pedagogically correct hints for four Common Core math standards at the fifth and sixth grade level.

Students must also learn data literacy to be computationally literate. If they do not understand how to work with data, they cannot effectively interact with the modern data-directed society. To improve data literacy instruction, we designed and tested an assignment built to help students understand the ways in which data can be misused and to help students recognize intentionally misleading figures in data.

Computational literacy also requires understanding and internalizing the ethical issues that can arise when designing, building, and using technology; many institutions are changing curricula to reflect this fact. Given that many students will not take an entire course devoted to ethics, we investigated the effects on and responses from students exposed to ethical considerations as part of a module within other courses at the 200 and 300 levels.

Generative Artificial Intelligence (GenAI) is broadly impacting and reshaping nearly every discipline across academia, the workplace, and society. Students must understand what it is and be aware of its significant challenges and potential harms. We evaluated and synthesized recent works to explain its possible harms and discussed how to help students understand that while using GenAI may offer personalized learning experiences, it may reduce critical thinking if used improperly.

Finally, while not unique to computer science, for students to be considered field-literate, they need to retain what they learn. To that end, we discuss the results of implementing mastery-based grading and a unique token system that promotes both academic and social well-being on student success and opinions in intermediate and advanced undergraduate courses.

Advisor:

Beverly Woolf