Modeling and Supporting Creativity During Collaborative STEM Activities Creativity is one of the most crucial resources to Americas success in the 21st century. However, being creative is challenging for many students, because it requires persevering through challenges, attacking a problem from multiple perspectives, and being flexible in ones perspective, to name a few. One possible way to bolster individual creativity is through collaboration: That some problems may be better solved by two people working together than by one working alone has long been recognized. However, it is still not clear which factors improve creativity during collaborative problem solving, or how to best support creativity in collaborative contexts through technological means. The proposed research will contribute to human-centered computing, advancing a novel technological approach that relies on machine learning techniques in general and Natural Language Processing (NLP) in particular to assess creativity as students work together. As our testbed, we will use the established Affective Meta Tutor (AMT) system, which is a software environment for system modeling tasks. Currently, AMT is designed as a single user system, and so we will extend it to support student interaction through a standard chat-based interface, and more importantly, enhance AMT with NLP capabilities to automatically infer student creativity during collaborative activities. Our main research questions include the following: (1) which factors influence moment-by-moment creativity during collaborative problem solving activities? (2) how NLP detect those factors? (3) How can an adaptive tutor use this information to create personalized interventions that will support creativity during collaborative tasks? As the first phase in this research, to enrich understanding of creativity during collaboration, we will run a series of studies involving students solving problems in pairs with AMT and analyze the corresponding data to identify factors that are relevant to creativity processes and outcomes. Specifically, for the creativity process data, we will hand code students chat and session transcripts for salient events of interest; creativity outcomes will be established using consensual assessment and established rubrics, extended as necessary. For the second phase, this data will be used to derive computational models for assessing student creativity in terms of both moment-to-moment processes and outcomes through machine learning methodologies focusing on an NLP approach. The models will be evaluated by comparing them to the gold standard of the hand coded data; in addition to providing automatic assessment, we expect that the models will also inform factors that influence creativity during collaboration through educational data mining techniques we will apply. As the final phase, we will rely on our experience from the second phase to pilot test a set of interventions to foster creativity as students are working together, tailored to their current progress, in order to maximize creative outcomes. The intellectual merit of this research includes: (1) Using data corresponding to pairs of students solving open ended STEM-based problems to develop a rich and nuanced understanding of creativity processes and outcomes in collaborative contexts, and how these relate to knowledge, affect and creative thinking styles; (2) relying on that understanding to develop and evaluate novel NLP-based models that recognize salient, creativity-related events during collaborative open ended STEM activities; (3) Developing and conducting preliminary evaluation of personalized support for creativity during collaborative activities. The broader impacts of the proposed research include: (1) experiments will be conducted in with undergraduate students at ASU, focusing on Computer Science, Informatics, and Engineering majors who are the appropriate population for our domain. We will involve 50% female students in the research, who will advance their creative skills in ways that have been demonstrated to increase interest and involvement in STEM courses and careers; (2) The PI serves on ASUs Program Committees for Informatics, charged with advancing new Informatics-based undergraduate and graduate programs, including core courses in human-centered computing and educational Informatics. Undergraduate and graduate courses, projects, and research experiences will be advanced through the integrated research and education agenda; (3) Although advances have been made in supporting creativity through various means, to the best of our knowledge none have focused on automatically assessing the process of creativity during collaborative problem solving. This project will pave the way for a new class of cyberlearning technologies to both assess and foster creativity, through just-in time personalized support.
The PI is currently funded by the NSF grant: Modeling and Supporting Creativity During Collaborative STEM Activities. This proposal seeks REU supplemental funding for two undergraduate students, with the following two goals: (1) to provide undergraduate students with the opportunity to gain experience with research through a set of hands-on, well defined tasks; (2) to enhance the above-stated project by refining the project applications feedback capabilities and evaluation of these facilities. The objective of the Modeling and Supporting Creativity During Collaborative STEM Activities project is to provide tailored support for creativity through two related thrusts. First, we will develop models of student creativity using data coming from students collaborative STEM activities, which will use as input features extracted with Natural Language Processing (NLP). In particular, we will have students interact using chat functionalities in order to solve open ended tasks, and use their chat data to automatically assess creativity through computational models we will develop. We will then rely on the model output to design adaptive interventions to encourage creativity during problem solving. To date, we have tested several open-ended activities through pilot evaluations in order to identify activities that are suitable for enabling creativity and collaboration, including an educational game called Newtons Playground, as well as more traditional tasks (e.g., hobbits and ogres river crossing task). This proposal extends the scope of our project through additional research and development tasks that are appropriate for undergraduate students these tasks are described below.
|Effective start/end date||8/1/13 → 8/31/14|
- NSF-EHR-DUE: Division of Undergraduate Science, Engineering, & Mathem: $92,052.00