Informs Annual Meeting 2017

SA47

INFORMS Houston – 2017

SA48

2 - Speedscaling in Flow Shops with Lot Streaming Kan Fang, Tianjin University, Tianjin, China, kfang@tju.edu.cn, Michael L.Pinedo The variable frequency drives on electrical motors have become more attractive in manufacturing to save their energy. We consider using speed scaling techniques to solve the flow shop lot streaming problems. We investigate the structural properties of optimal schedules and the computational complexity of various scheduling problems. 3 - Optimality of a Priority Policy for a Server Scheduling Problem with a Deteriorating Server Jefferson Huang, Cornell University, Ithaca, NY, 14853, United States, jefferson.huang@cornell.edu, Mark E. Lewis, Cheng-Hung Wu We consider a server scheduling problem motivated by the control of semiconductor manufacturing systems, where the service rate depends on both the class of the job and the degree to which the server has deteriorated. Conditions are given under which a priority policy is optimal. The case where the server may be preemptively repaired is also considered. 4 - Scheduling for Map-reduce Jobs Vaneet Aggarwal, Purdue University, 315 N. Grant St., West Lafayette, IN, 47906, United States, vaneet@purdue.edu, Tian Lan The Map-Reduce computing framework rose to prominence with datasets of such size that dozens of machines on a single cluster were needed for individual jobs. The interesting aspect of map-reduce job scheduling is that the reduce jobs can start when all the map tasks have finished. We give competitive scheduling algorithms for such job scheduling procedures where the release times of jobs depend on completion times of others. 360F Value of E-Businesses Sponsored: EBusiness Sponsored Session Chair: Qizhi Dai, Drexel University, Philadelphia, PA, 19104, United States, qd24@drexel.edu 1 - Crowdfunding Lender Team Structure and Lending Behavior Bin Wang, University of Texas Rio Grande Valley, Edinburg, TX, United States, bin.wang@utrgv.edu, Diego Escobari The proliferation of crowdfunding in the last few years has drawn academic researchers’ interest. Specifically, many have examined lending behavior and borrowers’ success of person-to-person (P2P) lending. The current research proposes to examine how lender team structure affect lending behavior and campaign success. Using data collected from Kiva.com, a leading P2P platform connecting lenders and borrowers from around the world, we examine how the network positions of lending team captains and members affect their lending behavior and the success of crowdfunding campaigns. 2 - Evaluating Observational Learning in Competitive Two-sided Crowdsourcing Market: A Bayesian Inferential Approach Emmanuel Ayaburi, PhD Candidate, The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX, 78249, United States, emmanuel.ayaburi@utsa.edu, Zhechao (Charles) Liu, Yoris A. Au This study examines the effect of observational learning by both buyers and creatives (suppliers) in the increasingly popular crowdsourcing market. Using a Bayesian framework, we estimate two structural models that capture buyers and creatives’ utility. Our results show that both buyers and creatives actively exercise observation learning from various market signals, and such learning leads to a more accurate prediction of market outcome and strong network effects. Our findings provide important insights on how crowdsourcing market participants can leverage the influence of the information signals present in the market to enhance their utility and promote a sustainable growth of the market. 3 - Crafting Personalized Incentives: A Randomized Field Experiment in an Online Dating Platform Ting Li, Erasmus University, T09-14, Burg Oudlaan 50, Rotterdam, 3000DR, Netherlands, tli@rsm.nl, Rodrigo Belo Heterogeneity across consumers often leads to diverse responses to incentives. We develop a method that accounts for consumer heterogeneity and maximizes consumer life-time value (CLTV), by personalizing incentives in scenarios of multiple conversion outcomes. We use data from a real-world randomized experiment in an exclusive online dating platform to create uplift models that predict changes in consumer behavior. We apply these methods in a new out-of- sample randomized experiment and find considerable improvements in CLTV, when compared to randomly assigned incentives. We explore the effectiveness of personalized incentives across consumers and discuss implications for business. SA47

361A Behavioral Issues in Supply Chain and Service Management Sponsored: Behavioral Operations Management Sponsored Session

Chair: Tingliang Huang, Carroll School of Management, Boston College, Chestnut Hill, MA, 02467, United States, tingliang.huang@bc.edu Co-Chair: Hang Ren, University College London, London, N12 8AQ, United Kingdom, hang.ren.13@ucl.ac.uk 1 - The Behavioral Promise and Pitfalls in Compensating Operations Managers Shan Li, Assistant Professor, Baruch College, City University of New York, 55 Lexington Avenue, New York, NY, 10010, United States, shan.li@baruch.cuny.edu, Kay-Yut Chen, Ying Rong We theoretically and behaviorally studied Profit-Sharing and Target-with-Bonus compensation schemes in an inventory management context. Our experimental data reveals behavioral promise and pitfalls of the two widely used incentive compensation schemes, and it further suggests systematic deviations from the theoretical benchmarks. Based on the experimental findings, we manipulate the behaviors and engineer the design of the compensation scheme through two additional treatments, to explore potential improvement of the mechanism. Our results have implications for the design of incentive schemes in practice. 2 - Service Systems with Unknown Quality and Customer Anecdotal Reasoning Hang Ren, UCL.School of Management, Flat 52, Nansen Village, 21 Woodside Avenue, London, N12 8RW, United Kingdom, hang.ren.13@ucl.ac.uk, Tingliang Huang We consider a service system where customers do not know the distribution of uncertain service quality and cannot estimate it fully rationally; instead, they form beliefs by taking the average of anecdotes. The sample size measures to what extent customers are boundedly rational. We characterize customers’ joining behavior and the server’s pricing, quality, and information disclosure decisions. As customers gather more anecdotes, the server may first decrease and then increase price, and the revenue is U-shaped. Interestingly, more anecdotes may induce a lower quality. Moreover, a low (high)-quality service provider may (not) disclose quality information if the sample size is small (large). 3 - Non-binding Goals in Teams: A Real-effort Coordination Experiment James Fan, juf187@smeal.psu.edu We experimentally investigate the efficacy of non-binding (wage-irrelevant) team goals. In our study, participants act as either team workers or managers. A manager can set a non-binding goal for the team production, which is determined by the minimum (or weak-link) performance of its workers. Consistent with our theory, we find evidence that the team production increases when managers are able to set goals. This effect is stronger when goals are challenging but attainable for weak-link workers. We also find evidence that many managers assign goals that are too challenging for weak-link workers, resulting in suboptimal team production and profits. 4 - Trust Among Executives Yanchong Zheng, Massachusetts Institute of Technology, Sloan School of Management, 100 Main Street, E62-578, Cambridge, MA, 02139, United States, yanchong@mit.edu, Emily Choi, Ozalp Ozer We integrate the results of a social network survey and a forecast information sharing experiment to determine whether and how trust and trustworthiness impact high-ranking executives’ decisions in a dyadic supply chain. The members of our executive sample have on average 17 years of work experience. Over half of them hold C-level positions in world-leading organizations. We demonstrate that executives are significantly motivated by trust and are more cognizant of when to rely on trust in their business decisions than non-experienced individuals. These results offer valuable insights into how organizations can better leverage executives’ trust intelligence to improve supply chain efficiency.

18

Made with FlippingBook flipbook maker