Academic success requires not only taking in content, but also understanding how to learn best. Self Regulated Learning (SRL) is process by which humans regulate their thinking, emotions, and behavior. It broadly describes the process of knowing (or learning) how to learn. Education research has found Self-Regulated Learning to be a key predictor of academic success along with other constructs like motivation and self-efficacy. It may be particularly critical in learning to program at the post-secondary level. Studies have shown that students benefit greatly from targeted instruction in these skills. Teaching students how to better self-regulate is both important and valuable for Computer Science students.
The solution here may seem straightforward: educators should give instruction on self-regulation skills. However, there are a number of skills that encompass a student's proficiency with self-regulate; including time management, problem decomposition, and reflection. Self regulation also tends to be a highly cognitive and internal process making it difficult to observe directly, let alone measure.
Which skills should be prioritized for targeted instruction? How could we empirically measure those skills? What limitations should we keep in mind when making such decisions? Within this dissertation, I will seek to address these questions. In order to get an idea of what skills the Computing Education Research community should be prioritizing, my co-authors and I conducted two studies. First, a Delphi Process study that expanded the field by gaining an understanding of what SRL skills CS post-secondary educators value most. This gave a more firm view of what skills were most important for CS students. Second, a systematic literature review to examine what skills had been studied within the Computing Education Research community. Ultimately, I created a finalized list of 12 SRL skills that appear to be particularly important to CS education. This list also includes behaviors an outside observer could use as indicators of the presence or absence of SRL.
After creating this list, I then considered how best to measure these each of these 12 skills. One form of measurement comes from using data traces collected from educational software. These allow researchers to make strong inferences about a student's internal state empirically. They also allow for measurement of students at greater scale and through automated means, making them advantageous for large classes. For my third publication, I then set about identifying a set of data traces for these skills taking a theory-first approach. I also make the case that CS is well situated to make great gains in trace-based approaches as they make use of a whole ecosystem of data sources. This is important as it is currently common for studies to utilize just one. / Doctor of Philosophy / Knowing how to learn is a critical aspect to academic success. Self-Regulation is the process by which humans regulate their thinking, emotions, and behavior. It encompasses the process of knowing (or learning) how to learn. Several studies have argued that learning Computer Science especially requires a strong self-regulated learning, but studies show novice programmer's skills in this area are still weak and benefit from further instruction. This is true even for students entering post-secondary education. Thus teaching students how to better self-regulate is important for CS students, but creating such lessons is not straightforward. SRL is a broad field and covers a variety of different skills that students may need. What skills are most important for instructors to teach their students? Once we know what skills are most important for targeting, how do we measure those skills? These are the questions I examine.
In order to get an idea of what skills the Computing Education Research community should be prioritizing, I conducted both a Delphi Process study. Following that I conducted a systematic literature review to get a better idea of what the Computing Education Research community is currently studying. I then considered the best way to measure these skills. While there are many approaches available to study SRL, I opted to examine these skills through student interactions with digital education software, called data traces. These traces are advantageous as they authentically capture learning in a way no other approach currently can. For my third paper I systematically derived a series of high-quality traces and made the case that CS classes already collect a lot of valuable traces through common digital education software systems.
| Identifer | oai:union.ndltd.org:VTETD/oai:vtechworks.lib.vt.edu:10919/120983 |
| Date | 21 August 2024 |
| Creators | Domino, Molly Rebecca |
| Contributors | Computer Science and#38; Applications, Shaffer, Clifford A., Jones, Brett D., Hooshangi, Sara, Edwards, Stephen H., Edmison, Kenneth Robert, Jamieson, Alan |
| Publisher | Virginia Tech |
| Source Sets | Virginia Tech Theses and Dissertation |
| Language | English |
| Detected Language | English |
| Type | Dissertation |
| Format | ETD, application/pdf |
| Rights | Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International, http://creativecommons.org/licenses/by-nc-sa/4.0/ |
Page generated in 0.0245 seconds