Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

SD14

n SD13 North Bldg 126B

n SD14 North Bldg 126C Data and Platforms Sponsored: Manufacturing & Service Oper Mgmt/Service Operations Sponsored Session Chair: Hamsa Sridhar Bastani, Wharton School, Philadelphia, PA, United States Co-Chair: Kenneth Moon, University of Pennsylvania, Philadelphia, PA, 19104, United States 1 - Network Effects in Contagion Processes Identification and Control Fanyin Zheng, Columbia University, Columbia Business School, 412 Uris Hall, New York, NY, 10027, United States, Kimon Drakopoulos In this paper, we study the problem of identifying network effects in contagion processes and present an application to the propagation of influenza in the United States. In particular, using data on the evolution of infections over time, the travel intensity between states as well as environmental conditions we first provide a framework to identify the true network effect of traveling between states. We use our estimates to propose and evaluate the performance of intervention and control policies, illustrating the benefits of network-based interventions. 2 - The Value of Pop-up Stores in Driving Online Engagement in Platform Retailing: Evidence from a Large-scale Field Experiment with Alibaba Dennis Zhang, University City, MO, 63124, United States, Hengchen Dai, Lingxiu Dong Short-lived and experiential-oriented pop-up stores have become a mainstream retail strategy. We provide the first causal evidence on how pop-up stores affect consumers’ subsequent behaviors. In a randomized field experiment involving approximately 800,000 consumers with Alibaba Group ù China’s largest e- commerce company ù we randomly assigned consumers to either receive a message about an upcoming pop-up store event or not receive any message. Our results show that visiting a pop-up store not only increases customers’ engagement with participating brands, but also improve customers’ long-term engagement with the platform. 3 - Optimal Recommendations and Preference Learning with Customer Disengagement Divya Singhvi, MIT, Hamsa Sridhar Bastani, Pavithra Harsha, Georgia Perakis We consider the problem of sequential product recommendations when customer preferences are unknown and customers are restless (i.e., customers decide to stay on the platform based on the quality of recommendations). In particular, we model customers who may abandon a content recommender (e.g., Pandora, Netflix) for an outside option (e.g., Apple Music or Hulu) based on the quality of recommendations they have received so far. We prove that bandit learning algorithms over-explore and the greedy policy under-explores in this regime. We propose a new learning algorithm that carefully balances the exploration- exploitation tradeoff in this setting. 4 - Flexibility and Relationships in Online Marketplaces Jiding Zhang, The Wharton School, 3730 Walnut St, 500 Jon M. Huntsman Hall, Philadelphia, PA, 19104, United States, Elena Belavina, Karan Girotra, Kenneth Moon Online marketplaces have grown and diversified as intermediaries for services, extending from the flexible (Uber and Airbnb) to those in which trust and relationships are valued ($800B home services market). Using data from a leading online labor marketplace, we empirically relate the operational performance of the platform’s pricing and matching regimes to a simple market characteristic: how clients derive comparative value from flexibility and/or relationships with service providers. Extending recent theoretical modeling in dynamic mechanism design, our structural empirical methods use large-scale market data to prescribe pricing and matching policies. 5 - Adaptive Learning with Unknown Information Flows Ahmadreza Momeni, Stanford University, Yonatan Gur We introduce a generalized multi-armed bandit formulation in which additional information on each arm may appear arbitrarily throughout the decision horizon, and study the impact of such information flows on the achievable performance and the design of efficient decision-making policies. We characterize the regret complexity of this family of problems and introduce a general and practical adaptive exploration approach for designing policies that, without any prior knowledge on the information arrival process, attain the best performance that is achievable when the information arrival process is a priori known.

Navigating the Publication Process: Recently Published Papers in Empirical Healthcare OM Sponsored: Manufacturing & Service Oper Mgmt/Healthcare Operations Sponsored Session Chair: Song-Hee Kim, University of Southern California, Los Angeles, CA, 90089, United States 1 - When Waiting to See a Doctor is Less Irritating: Understanding Patient Preferences and Choice Behavior in Appointment Scheduling Nan Liu, Boston College, 140 Commonwealth Avenue, Fulton Hall, Room 340, Chestnut Hill, MA, 02467, United States, Stacey Finkelstein, Margaret Kruk, David Rosenthal In this talk, we share the journey of our recent research and its publication process. This study concerns patient preferences and choice behavior in scheduling medical appointments. We conduct four discrete choice experiments on two distinct populations and identify several “operational attributes (e.g., delay to care and choice of doctor) that affect patient choice. We observe an interesting gender effect with respect to how patients tradeoff speed (delay to care) vs. quality (doctor of choice), and demonstrate that risk-attitude is a mediator variable. Our results have important implications for improving patient experience in a medical practice based on patient mix and current delay level. 2 - Variety and Experience: Learning and Forgetting in the Use of Surgical Devices Kamalini Ramdas, London Business School, A215 Sussex Place, Regent’s Park, London, NW1 4SA, United Kingdom We examine learning and forgetting in hip surgery as a function of surgeons’ experience with specific device versions. We develop a generalizable method to correct for left censoring of granular experience data, using maximum simulated likelihood estimation. With steep learning and rapid forgetting, device variety hurts surgeons’ productivity and quality. 3 - Team Familiarity and Productivity in Cardiac Surgery Operations: The Effect of Dispersion, Bottlenecks, and Task Complexity Bilal Gokpinar, UCL School of Management, 1 Canada Square (38th floor), Canary Wharf, London, E14 5AA, United Kingdom, Emmanouil Avgerinos Fluid teams are commonly used by a variety of organizations to perform similar and repetitive yet highly critical and knowledge-intensive tasks. Using a granular data set of 6,206 cardiac surgeries from a private hospital in Europe over seven years, our study offers a new and detailed account of how shared work experience influences team productivity. We highlight the role of nuanced team composition dynamics beyond average team familiarity. We observe that teams with high dispersion of pairwise familiarity exhibit lower team productivity, and the existence of a “bottleneck-pair may significantly hinder overall knowledge transfer capability, thus, productivity of fluid teams. 4 - The Impact of E-Visits on Visit Frequencies and Patient Health: Evidence from Primary Care Hessam Bavafa, Assistant Professor, Wisconsin School of Business, 975 University Avenue, Madison, WI, 53706, United States, Lorin Hitt, Christian Terwiesch Secure messaging, or “e-visits,” between patients and providers has increased sharply in recent years, and many hope they will help improve healthcare quality while increasing provider capacity. Using a panel dataset from a large healthcare system in the United States, we find that e-visits trigger about 6% additional office visits, with mixed results on phone visits and patient health. These additional visits come at the sacrifice of new patients: physicians accept 15% fewer new patients each month following e-visit adoption. Our dataset on nearly 100,000 patients spans from 2008 to 2013, which includes the rollout and diffusion of e-visits in the health system we study.

95