Drs. Pete Vanden Bosch and Jinhyok Heo
Judges choice first place poster winner, Dr. Pete Vanden Bosch
Dr. Pete Vanden Bosch works for the Institute for Defense Analysis and Marymount University. He retired from USAF in 2010 as the chief analyst for NORAD and U.S. Northern Command. His current research interests are in optimization and the psychology of decision-making.
The Goldilocks heuristic is the tendency of decision makers, when presented with a range of options, to choose an intermediate one. In many cases, this heuristic works very well. We drive in the middle of our lane without thinking about the potential risks of hugging one side. We discount outliers in data sets. We seek political compromise.
But when we apply a heuristic to inappropriate situations, it becomes a fallacy. In particular, the Goldilocks heuristic is inappropriate to multicriteria decision-making, and using it to make such decisions is worse than mere guessing. Nevertheless, psychosocial forces driving us to use it are compelling. Pete’s contribution was to provide examples from DoD and DHS, quantify the fallacy, and connect the situation to psychosocial drivers like anchoring and prospect theory.
Judges choice third place winner and members choice 2nd place winner, Dr. Jinhyok Heo
At the SRA meeting last December, I was surprised two times when the best poster awards were announced; my poster was selected not only by the conference judges (for the third place) and but also by the participant votes (for the second place)! I presented a new model called “The Air Pollution Social Cost Accounting (APSCA) model,” which I developed at Cornell University as a postdoc since last March, right after I had finished my doctoral study at Engineering and Public Policy (EPP) in Carnegie Mellon University. In my opinion my work has been acknowledged for two reasons: scientific achievement and clear presentation.
The APSCA model identifies the sources of air pollution affecting a certain location (or receptor) and quantifies their contributions with unprecedented detail and efficiency. Although it may sound basic, such accounting has been a challenge in air quality research because there exist innumerable responsible emission sources and air pollutants travel long distance (e.g. hundreds of kilometers) while they undergo complex chemical reactions. The APSCA model is built based on a model called “The Estimating Air pollution Social Impact Using Regression (EASIUR, available at <http://barney.ce.cmu.edu/~jinhyok/easiur/>) model,” a major part of my PhD dissertation at Carnegie Mellon. Derived from running regressions on a large dataset generated by a computationally demanding state-of-the-art air quality model, the EASIUR model estimates the sum of public health costs or social costs of emissions imposed on large regions surrounding any emission source located in the United States. I found out an efficient method to allocate the EASIUR’s social cost estimates to affected locations. By doing the spatial allocations for all the sources in the national emissions inventory, I built the APSCA model. Evaluations show that both EASIUR and APSCA produce estimates comparable to their parent state-of-the-art model but without costly computations.
In addition to what they can already contribute to policy research community, I am very excited about future works. First of all, I am currently working on combining the two models with optimization methods to build another model that will assist in optimal decision-making associated with air quality, energy, and climate change. Next, the two models will be able to keep linking up-to-date science with policy research because they can be updated as new understanding or data (e.g. emissions and meteorology) are fed to their parent state-of-the-art air quality model. Lastly, I plan to derive models for other important regions such as China, India, and Europe. Along the way, my models would open up many other opportunities for original research.
I believe the art of preparing a good poster or presentation lies in creating a pleasant tension between content and audience. Because my audience are often busy and from multiple disciplines, I always try my best to present my research succinctly using plain language, intuitive figures and diagrams, and proper colors and fonts. I learned a lot from my own trial and errors as well as good examples during my PhD study at the strong interdisciplinary research environment of Carnegie Mellon’s EPP. I’d like to thank my co-authors: my PhD advisor, Prof. Peter Adams, for his superb guidance on all aspects of my academic life and my postdoc advisor, Prof. Oliver Gao, for his generous support for pursuing my research ideas.
Dr. Jinhyok Heo (firstname.lastname@example.org)