This SBIR Phase I project proposes to build a Cyberlearning tool called REACT (Real Time Evaluation and Assessment of Computational Thinking) based on a highly innovative assessment methodology that will help teachers with computer science education. There is growing interest in having more students exposed to computer science and in broadening STEM participation for women and minority students. However, the question is one of how. Research by the REACT team - conducted in diverse innercity, remote rural and Native American communities - indicates that: a) a well-defined strategy exists that can make computer science education work in public schools, and b) it is difficult to create the classroom environment needed to support the highly engaging and educational pedagogical approaches that form the heart of this strategy. Open ended projects can be highly inspiring to students across genders and ethnic groups. It is proposed to build and evaluate a cyberlearning tool that will assess student-built games (or simulations) and provide the teacher with feedback on what students are actually learning in real time. This feedback is an embedded, formative assessment that gives the teacher important information directly from their students' programs and enables teachers to make crucial instructional decisions during lessons. The broader/commercial impact includes the benefits of embedded assessment - that is, assessment that taps directly into student created projects instead of soliciting information from students through tests. Part of the assessment problem is that it is difficult to extract meaningful information from design artifacts. In the case of game design, it is most useful to extract information from the programs and games themselves. REACT employs a highly innovative pattern matching approach called Computational Thinking Pattern Analysis (CTPA). CTPA can find patterns that are relevant to game design as well as computational science modeling. A successful embedded, formative assessment of design-based artifacts like programs could advance learning sciences and significantly enhance in-class and online learning by creating closed loop learning environments presenting real time information to students and teachers. REACT could result in an important tipping point because embedded, formative assessment provides crucial information to teachers as they are making crucial instructional decisions. This could be highly relevant to broadening participation in computer science learning. A teacher using a supportive Cyberlearning approach would be able to better engage and motivate female and underserved populations. This increased engagement could significantly scale up the exposure of students to computer science education.
