Skill sets such as understanding and applying computational concepts are essential prerequisites for success in the 21st century. One can learn computational concepts by taking a traditional course offered in a school or by self-guided learning through an online platform. Collaborative learning has emerged as an approach that researchers have found to be generally applicable and effective for teaching computational concepts. Rather than learning individually, collaboration can help reduce the anxiety level of learners, improve understanding and create a positive atmosphere to learning Computational Thinking (CT). There is, however, limited research focusing on how natural collaborative interactions among learners manifest during learning of computational concepts. Structured as a manuscript style dissertation, this doctoral study investigates three different but related aspects of novice learners collaboratively learning CT. The first manuscript (qualitative study) provides an overall understanding of the contextual factors and characterizes collaborative aspects of learning in a CT face-to-face classroom at a large Southeastern University. The second manuscript (qualitative study) investigates the social interaction occurring between group members of the same classroom. And the third manuscript (quantitative study) focuses on the relationship between different social interactions initiated by users and learning of CT in an online learning platform Scratch™. In the two diverse settings, Chi's (2009) Differentiated Overt Learning Activities (DOLA) has been used as a lens to better understand the significance of social interactions in terms of being active, constructive and interactive. Together, the findings of this dissertation study contribute to the limited body of CT research by providing insight on novice learner's attitude towards learning CT, collaborative moments of learning CT, and the differences in relationship between social interactions and learning CT. The identification of collaborative attributes of CT is expected to help educators in designing learning activities that facilitate such interactions within group of learners and look out for traits of such activities to assess CT in both classroom and online settings. / PHD / One of the overarching processes defining the future is the digital revolution, impinging on, reshaping, and transforming our personal and social lives. Computation is at the core of this change and is transforming how problems are defined, and solutions are found and implemented. Computer modeling, simulation and visualization software, Smart grid, and Software Defined Radio, are few examples where computation has allowed us to tackle problems from varied perspectives. Vast domains await discovery and mapping through creative processes of Computational Thinking (CT). CT is the thought process that enables us to effectively work in such a technology driven collaborative society. It provides us the ability to find the right technology for a problem and apply technology to resolve the problem. Skill sets such as understanding and applying computational concepts are essential prerequisites for success in the 21st century. One can learn CT by taking a traditional course offered in a school or by self-guided learning through an online platform. This doctoral study investigates three different but related aspects of how new learners are learning CT. The first qualitative study provides an overall understanding of circumstantial factors that influence the learning in a CT face-to-face classroom at a large Southeastern University. The second qualitative study investigates how students in groups (in the same classroom setting) can help each other to learn CT. And the third quantitative study focuses on users’ learning of CT in an online learning platform Scratch™. Together, the findings of this dissertation study contribute to the limited body of CT research by providing insight on new learner’s attitude towards learning CT, collaborative moments of learning CT, and the differences in the relationship between social interactions and learning CT. The identification of collaborative attributes of CT is expected to help educators in designing learning activities that facilitate such interactions within a group of learners and look out for traits of such activities to assess CT in both classroom and online settings.

Self-regulated learning skills have been shown to positively impact academic achievement in educational settings. This same set of skills becomes critically important as graduates enter today‟s dynamic work environment. That environment increasingly requires accountants and other professionals to be lifelong learners. This study is a response to the call of the Accounting Education Change Commission (AECC) to make "learning to learn" a priority in the accounting classroom. This study used a quantitative, quasi-experimental design within the context of a beginning accounting course. The course is characterized by high failure rates, highly conceptual content, and a population of novice learners. Study participants were stratified according to ACT level, prior GPA level, and academic major. The control group received instruction based on course content only. The treatment group received an intervention in which instruction focused on the process of learning as well as on regular course content. The purpose of the study was to determine whether academic performance differed between the 2 groups. The study further examined whether differences in the means on exam scores varied as a function of ACT level, prior GPA level, or academic major. A MANOVA indicated a significant difference in exam scores between the control and treatment groups with the treatment group outperforming the control group on 4 of the 5 exams. Follow-up ANOVAS were used to determine on which exams statistical significance was found. Two-way ANOVAS revealed no significant interaction between classroom method and prior GPA or academic major. Although statistical significance was not found in the interaction between classroom method and ACT level, descriptive statistics revealed that the greatest differences (between the mean exam scores of the control and treatment groups) occurred among the high-ACT group. The researcher did not attempt to trace causal paths, or changes in the mediating variables that may have linked changes in the learning environment to changes in academic performance. This study provided evidence that instruction related to the process of learning did not diminish academic performance on content-based exams and for most groups of students who received such instruction performance on exams was enhanced. This study casts doubt on the presence of a "ceiling effect," which is often associated with high-ACT students.