آزمایش انسانی در خصوص تصمیمات موجودی با توجه به عدم قطعیت عرضه
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی|
|20753||2013||13 صفحه PDF||33 صفحه WORD|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : International Journal of Production Economics, Volume 142, Issue 1, March 2013, Pages 61–73
2.عوامل بررسی شده و فرضیات تست شده
2.1بازی آبجو و تاثیر شلاق چرمی با زمان های فرآوری تصادفی
2.2عدم قطعیت عرضه و تصمیمات موجودی: یک دیدگاه رفتاری
جدول 1. شرکت کنندگان در روش های مختلف.
4.شبیه سازی عددی سناریوهای بازی آبجو
شکل 1. انحرافات معیار سفارشات بازیگر مجازی.
شکل 2. هزینه های کل موجودی بازیگر مجازی.
5.نتایج آزمایشات انسانی
5.1شواهد تاثیر شلاق چرمی
شکل 3. سفارشات انباشته بازیگر مجازی براساس دوره و نقش.
جدول 2. انحراف معیار میانگین و میانه، براساس نقش و روش.
جدول 3. آزمون ویلکاکسون در خصوص گسترش انحرافات معیار، و روش.
5.2سفارشات، مالکیت موجودی و عملکرد SC
شکل 4. سفارشات انباشته تجربی میانه براساس نقش و روش.
شکل 5. هزینه های انباشت تجربی میانه براساس نقش و دوره.
شکل 6. هزینه های میانه مالکیت موجودی، انباشت و کل هزینه های موجودی براساس نقش.
جدول 4الف. پانل پویا- خرده فروش.
جدول 4ب. پانل پویا- عمده فروش.
جدول 4ج. پانل پویا- توزیع کننده.
جدول 4د. پانل پویا- کارخانه.
6.بحث و بررسی
نتیجه 1- تاثیر شلاق چرمی با توجه به عدم قطعیت عرضه افزایش می یابد.
نتیجه 2- افراد بشر در بازی آبجو به عدم قطعیت بزرگ تر با کاهش سفارشات
Controlled human experiments are adopted in this paper to investigate the impact of supply uncertainty on buyers' inventory management. The experiments aim at assessing the impact of one specific source of supply chain uncertainty, namely stochastic lead times, on inventory holdings and the extent of the bullwhip effect. Three experimental treatments are run within the framework of the beer game manipulating variability in demand and in lead times. Results confirm that the bullwhip effect arises in all experimental treatments and that the variance of orders is higher under stochastic lead times. Analysis of players' behaviour in the course of the game suggests that players react to higher uncertainty by holding fewer inventories, a behaviour consistent with the predictions of some psychological models of choice under ambiguity.
An increasing degree of complexity characterises the supply chains of many sectors (Esposito and Passaro, 2009). Among the causes of such a trend are outsourcing, the enlargement of supplier networks, increased dependence on supplier capabilities, shorter product life-cycles, and international market and production expansion (Wagner and Neshat, 2010). Further, as firms try to reduce costs through the rationalisation and reduction of the supply base, the aim to secure a stable flow of materials has become more difficult to achieve (Harland et al., 2003). As a consequence of both higher complexity and leaner supply chains, the instability of supply chains and supply uncertainty have increased (Geary et al., 2006). A paradigmatic representation of supply chain instability is the Bullwhip Effect (BWE). The BWE is generally triggered by demand uncertainty (Forrester, 1958), and it entails that, as external demand passes through the SC from the downstream to the upper levels of the chain, the variance of orders is amplified. This behaviour can imply substantial costs in terms of stock-out as well as inventory holding and obsolescence costs, thus worsening the performance of the SC. While the impact of demand variability on SC instability and performance has been explored in several studies (Croson and Donohue, 2006, Steckel et al., 2004, Gupta et al., 2002 and Sterman, 1989), the impact deriving from supply-side sources of uncertainty has received less attention, in spite of the fact that some authors have posited that a reduction in SC instability is best enabled via implementation of the principles of smooth material flow, and by decreasing actual or perceived shortage risk (Geary et al., 2002 and Geary et al., 2006). A small number of numerical simulations (Chatfield et al., 2004 and Truong et al., 2008) has investigated the effects of supply uncertainty on the extent and consequences of SC instability by making supply uncertainty operational through stochastic lead times, one of the most relevant supply-side sources of uncertainty. These studies have shown that, generally, stochastic lead times contribute to worsen SC instability. While numerical simulations can throw light on how rational and optimising agents can react to SC uncertainty, they cannot fully account for deviations from rationality or limited cognitive abilities of SC managers. In this direction, experimental research on human subjects in neighbouring disciplines to Operations Management has shown that decision makers apply heuristics in processing tasks characterised by uncertainty (Kahneman and Tversky, 1974), and that they may use these heuristics as a way “to live with risk” (Gigerenzer, 2002). Further, decision makers exhibit biases in processing probabilistic information, since they distort probabilities of outcomes even when they are objectively known (Kahneman and Tversky, 1979) and, according to the domain of outcomes (costs vs. revenues), they might dislike/prefer uncertainty (Ellsberg, 1961 and Wakker, 2010). Controlled human experiments have gained importance as a methodology for the study of SC instability since Sterman's (1989) finding that the BWE is a problem arising as a consequence of human decision making and stemming from the amplification of unanticipated changes in demand, and from a biased perception of the flows in transit through the SC pipeline. In the face of varying types and degrees of SC uncertainty, managers' risk mitigation actions depend on their individual attitudes, on their perceptions of the likelihood of supply disruptions, and on the degree of confidence they assign to available risk information (Ellis et al., 2010). Thus, there are grounds for positing that under supply uncertainty heuristics and biases used in assessing probabilities of outcomes affect either the size of the BWE, or inventory holdings and inventory policies, or both, in ways which might be at odds with the predictions of numerical simulations. Further, since heuristics may develop through time in repeated tasks, it is relevant to provide insight into the way individuals adapt their decisions as they experience a highly variable environment. This analysis can hardly be carried out in the field because of lack of sufficient control, thus suggesting that human experimentation may prove a useful tool to explore the effects of supply uncertainty on the BWE formation and SC costs (Bendoly et al., 2006 and Ancarani and Di Mauro, 2011). To the best of our knowledge, a test of the impact of supply uncertainty on SC performance using human subjects is still lacking. In this paper, we apply human experiments with the aim to study the behaviour of members of a SC in the face of lead time uncertainty, and contrast the performance of a SC with stochastic lead times with that of a SC with deterministic lead times. We carry out the study within the framework of the beer distribution game, which reproduces a serial supply chain with four echelons (retailer, wholesaler, distributor, and factory). Two research questions are investigated: 1. What is the impact of supply uncertainty on supply chain instability, as measured by the BWE, and on supply chain performance, as measured by SC costs? 2. How do SC managers react to supply uncertainty in terms of ordering decisions and inventory holdings? Our results show that supply uncertainty, in terms of stochastic lead times, gives rise to a higher variance of orders at every echelon of the supply chain. More intriguingly, we find that, when the SC is characterised by both demand uncertainty and stochastic lead-times, buyers hold fewer inventories, a behaviour that we attribute to an uncertainty loving attitude. The paper is organised as follows: Section 2 reviews the relevant branches of the literature underpinning the hypotheses tested through the experiment, Section 3 presents the experimental design, while Section 4 develops benchmarks for behaviour in the experiment by means of numerical simulation. Section 5 presents the results of the human experiment. Section 6 discusses the main findings, and highlights implications for SC management and future research. Section 7 concludes the paper.
نتیجه گیری انگلیسی
Supply uncertainty is a major issue for inventory management in serial supply chains and lead-time uncertainty is one of the most prominent and common of its components. A potentially fruitful approach to this issue is to bring together the results and methodologies of two separate streams of literature. The first comprises behavioural studies that have long demonstrated that cognitive biases are pervasive in choice under uncertainty and have identified systematic deviations such as aversion or preference for ambiguity. The second comprises numerical simulation studies of the expected impact of supply uncertainty (and lead time uncertainty in particular). This paper has attempted to apply this approach and has investigated the impact of uncertainty in lead times on the formation of the BWE, inventory holding and SC performance through a controlled human experiment, whose results have been compared with those of a numerical model simulating a risk neutral, ambiguity neutral player. The paper positions itself among the contributions that aim at studying the relevance that SC vulnerability plays in SC management: in the experiment, lead time uncertainty in the presence of a single supplier can be considered a proxy of the vulnerability of the SC to stock-outs. Although the experiment has not been carried out with professional managers, the fact that participants were graduate students with background in Operations Management makes it reasonable to assume that these results may be descriptive of the behaviour of actual purchasing managers. The following limitations of the present study should be underlined: uncertainty in supply has been made operational through a specific distribution of lead-time, thus it is possible that different levels of uncertainty or different distributions give rise to different behaviour. In this vein, an interesting extension of the research may consider introducing stochastic lead times at only one echelon rather than at all four layers. Next, the possibility to link inventory holding to the individual ambiguity attitude has been restricted by the choice to manipulate the presence/absence of stochastic lead times on a between subject basis. However, the indubitable gain of the between subject manipulation is that the observed differences between players’ behaviour across treatments are free from potential framing effects created by the comparative setting. As part of our future research agenda we plan to address these shortcomings and to gain sound support for the effect of various facets of supply risk on SC performance.