Poster Presentation@CMU’s Teaching & Learning Summit
Peter Scupelli and Paul Inventado presented posters at the first Teaching and Learning Summit for faculty and graduate students hosted by Carnegie Mellon University’s Eberly Center for Teaching Excellence and Educational Innovation.
The event was designed to:
- Foster dialogue, networking and collaboration within and across disciplines.
- Showcase the educational research of CMU instructors and learning scientists.
- Share transferable, evidence-based and innovative teaching strategies used by CMU instructors.
1. Scupelli, P. and Brooks, J. “Dexign Futures: a flipped, open learning initiative course”
Design for sustainability opportunities reside in bridging between short-term action and long-term strategic thinking.
Unfortunately, traditional design pedagogy poorly equips designers for long-term strategic thinking. In the Dexign Futures class described in
this poster, students learn to align short-term design with long-term horizons. Dexign Futures is a required design studies class for all
third year undergraduate students in the products, communications, and environments tracks in the School of Design at Carnegie Mellon
University. Flipped courses shift lectures and instruction to the online learning initiative (OLI) course to use class time for hands-on
activities. Online homework helps students to prepare for in-class activities. During in-class activities, the course instructor, and
teaching assistants can provide students with feedback and answer questions. Likewise, in-class team activities and peer feedback can
enhance student learning. Research from piloting of the online modules and in-class workshops are promising. We measure student learning for this flipped class.
2. Inventado, P. S. and Scupelli, P. “A Data-Driven Design Pattern Methodology to Facilitate Effective Pedagogical Practice in Online Learning Systems”
Paradoxically, online learning system designs often fail to fully benefit from research insights and findings expressed as
learning principles possibly because of nuances in translating to specific learning contexts. We present a methodology that uses data
exploration, statistical analyses, and machine learning to uncover relationships between designs and learning outcomes. In a specific
context, effective designs linked to good outcomes are encapsulated into design patterns. For example, analyses of student interaction
data from an online learning system revealed that struggling students quickly learned to solve similar math problems by requesting all
available hints and answers, and learning from it. This was encapsulated into the Explain Worked Solutions design pattern:
incorporate worked-out-solutions in on-demand hints for students struggling with math problems. Design patterns are uncovered through
research, and their continued evaluation and refinement ensures reproducibility in tested contexts. Uncovered design patterns are currently compiled into an online repository to facilitate use.