Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

WC54

3 - The Impact of Service Level on Inventory Decisions Jaime Andres Castañeda, Universidad del Rosario, Bogota, Colombia, Sebastian Villa This study analyzes how a service level concern impacts inventory decisions. We experimentally manipulate decision makers’ mindset through priming techniques to (unconsciously) activate a service level mindset aligned with profit maximization. We also explicitly model a service level target in the Newsvendor problem, providing the analytical solution, model insights and suggestions for a behavioral test of this (consciously) activated service level mindset. 4 - Impact of Behavioral Factors on Performance of Multi-server Queuing Systems Marilyn T. Lucas, University of Vermont, School of Business Management, 202 Kalkin Hall 55 Colchester Avenue, Burlington, VT, 05405-0157, United States, Hung Do, Masha Shunko, David Novak Recent studies have shown that employees in service-based queuing systems respond to the design and congestion level of the queuing system in which they operate. In this paper, we incorporate two behavioral effects into multi-server analytical queueing models: server speedup due to increased workload and server slowdown due to social loafing when multiple workers share the workload. We evaluate how these factors affect the performance of both the multi-server, single- queue (or SQ) system and the multi-server, parallel-queue (or PQ) system. We also consider strategic routing and its impact on the performance of PQ systems. 5 - Construal Level Theory and the Newsvendor Problem Samuel Nathan Kirshner, UNSW Business School, Level 2, West Wing, Quadrangle Building, UNSW, Sydney, 2052, Australia, Chun-Tang Pai Experimental studies on the newsvendor problem has robustly observed that subjects choose order quantities that systematically deviate from the profit maximizing order quantity. Construal Level Theory (CLT) shows that people can be subconsciously primed by their environment to think more abstractly (high construal) or concretely (low construal), which in turn, can predictably influence decision-making. This talk investigates the potential impacts of CLT on demand chasing in the newsvendor problem. n WC54 North Bldg 232B Behavioral Operations Management Session Sponsored: Behavioral Operations Management Sponsored Session Chair: Xiaobo Zhao, Tsinghua University, Beijing, 100084, China Co-Chair: Wanshan Zhu, Tsinghua University, Beijing, 100084, China 1 - What Can We Learn from Brain Imaging Newsvendors? Kay-Yut Chen, University of Texas at Arlington, 511 Pine Island Circle, Mansfield, TX, 76063, United States The current literature on newsvendor behavior relies on “guessing the underlying cognitive mechanisms from observations, without direct physical evidence. Functional near infrared spectroscopy allows researchers to examine brain functions by measuring hemodynamic responses. In this paper, we show that newsvendor performances are correlated with neural activations, which can be interpreted as cognitive efforts. 2 - Full Credit or Full Quantity? A Behavioral Investigation of Buyback Contracts Yinghao Zhang, Assistant Professor, University of Cincinnati, 614A Carl H. Lindner Hall, 2925 Campus Green Dr., Cincinnati, OH, 45221, United States, Peiwen Yu, Tianjun Feng, Stephanie N. Eckerd This paper investigates two forms of buyback contracts: full quantity and full credit contracts. Standard theory suggests that the retailer should order the same amount of inventory under these two options. Lab experiments, however, reveal different ordering patterns. We propose a behavioral model to explain these experimental observations. 3 - How to Provide Information for Management Decisions Jaime Andres Castañeda, Universidad del Rosario, Bogotá, Colombia, Yingshuai Zhao This work experimentally studies several ways of providing information in a newsvendor task to pinpoint decision-making improvement interventions. Based on sunk-cost effects arguments, we manipulate (i) whether decision makers obtain supporting information for free or at a cost, and based on instrumental information arguments, we manipulate (ii) the feedback information provided. Results show (i) performance improvements when participants have to pay for supporting information and (ii) less order variability when they are provided with instrumental information.

4 - On Difference between Direct-response and Strategy Procedures in Decision Making: Neural Evidence in a Reward-punishment Game Niannian Sun, Tsinghua University, Beijing, China, Xiaobo Zhao, Lihong Wang, Yukun Zhao, Xue Xiao In decision making, the evidence of whether direct-response procedure and strategy procedure lead to similar outcome is mixed. We conduct an fMRI experiment of a game setting with reward and punishment decisions under both procedures. The results show that, in reward decision, the direct-response procedure prompts stronger neural unwillingness and lower reward rate than the strategy procedure. In punishment decision, the strategy procedure induces stronger neural negative emotions and higher punishment rate than the direct- response procedure. Moreover, subjects prefer not to choose reward when facing favorable states but tend to choose punishment when facing unfavorable states. n WC55 North Bldg 232C Product and Project Development and Management 1 Contributed Session Chair: Barbaros Yet, Hacettepe University, Ankara, Turkey 1 - Factors Affecting Labor Productivity in Turnaround Maintenance Projects: Case of the Oil and Gas Sector in Qatar Omar Ben-Ayed, Professor of Management, Qatar University, College of Business and Economics, Department of Management and Marketing, Doha, Qatar, Abdulhadi Nasser Al-Marri, Salem Nechi The aim of this study is to define the factors affecting labor productivity in Turnaround Maintenance (TAM) projects in the oil and gas sector in Qatar and to propose solutions to improve it. We use the Analytic Hierarchy Process method to determine the importance of the factors identified and to select the most appropriate managerial alternative for improvement. The target participants are mostly from the management level of TAM companies. The results show that the top five factors are labor skills, extent of supervision, communication skills, safety related issues, and on-site labor transportation. The study suggests improving the estimation of activity time and resources. 2 - Coordination between Iterations and Design Rate in Serial Product Development Process Maoqi Liu, PhD Candidate, Tsinghua University, Beijing, 100084, China, Li Zheng, Changchun Liu Product Development (PD) plays an important role in winning market competition. As complexity and uncertainty increasing in PD process, iteration becomes more and more significant in this process. Lead time is a key indicator to measure the performance of PD process. Many techniques like CPM and PERT are applied to non-iterative process but few of them can be applied in iterative one. In this research, a model is proposed to achieve best coordination between faster and fewer iteration with the object of minimizing lead time of PD process. The results in small scale shows that task with higher risk tends to design slowly to reduce iterations and therefore, shorten PD lead time. 3 - Probabilistic Earned Value Management using Bayesian Networks Barbaros Yet, Dr., Hacettepe University, Ankara, Turkey, Yasemin S Accurate project control is challenging due to uncertainty regarding the completion percentage of ongoing activities and uncertainty of the project plan. Our aim is to develop a probabilistic project control tool based on Earned Value Management (EVM). We use Hybrid Bayesian Networks (HBN) to model the causal relations between project risk factors and progress, and to reflect parameter uncertainty in EVM. The proposed tool is illustrated using a case study of a construction project.

488