1. Cai, Hongbin and Joseph Tao-yi Wang (2006), Overcommunication in Strategic Information Transmission Games, Games and Economic Behavior, 56(1), 7-36. (longer working paper version)
Abstract: In this paper we conduct laboratory experiments to test the Crawford and Sobel (1982) theory of strategic information transmission. Our experimental results strongly support the basic insight of the theory, namely, that less information is transmitted when preferences of the sender and the receiver diverge. Moreover, the average payoffs for the senders, the receivers, and the overall subject population are very close to those predicted by the most informative equilibrium. However, the evidence shows that subjects consistently overcommunicate in that the senders' messages are more informative about the true states of the world and that the receivers rely more on the senders' messages in choosing actions, compared with what the theory allows in the most informative equilibrium. To understand the overcommunication phenomenon, we use two popular approaches of bounded rationality: behavior type analysis and quantal response equilibrium, to analyze subjects' behavior in our experiment data.
2. Knoepfle, Daniel, Joseph Tao-yi Wang and Colin F. Camerer (2009), Studying Learning in Games Using Eye-Tracking, Journal of the European Economic Association, 7(2-3), 388-398. (long version with appendix)
Abstract: We report results from an exploratory study using eye-tracking recording of information acquisition by players in a game theoretic learning paradigm. Eye-tracking is used to observe what information subjects look at in 4x4 normal-form games; the eye-tracking results favor sophisticated learning over adaptive learning and lend support to anticipatory or sophisticated models of learning in which subjects look at payoffs of other players to anticipate what those players might do. The decision data, however, are poorly fit by the simple anticipatory models we examine. We discuss how eye-tracking studies of information acquisition can fit into research agenda seeking to understand complex strategic behavior and consider methodological issues that must be addressed in order to maximize their potential.
3. Kang, Min Jeong, Ming Hsu, Ian M. Krajbich, George Loewenstein, Samuel M. McClure, Joseph Tao-yi Wang and Colin F. Camerer (2009), The Wick in the Candle of Learning: Epistemic Curiosity Activates Reward Circuitry and Enhances Memory, Psychological Science, 20(8), 963-973. (2009 JCR ranking 6/111 in Psychology, Multidisciplinary; IF=5.09).
Abstract: Curiosity has been described as a desire for learning and knowledge, but its underlying mechanisms are not well understood. We scanned subjects with functional magnetic resonance imaging while they read trivia questions. The level of curiosity when reading questions was correlated with activity in caudate regions previously suggested to be involved in anticipated reward. This finding led to a behavioral study, which showed that subjects spent more scarce resources (either limited tokens or waiting time) to find out answers when they were more curious. The functional imaging also showed that curiosity increased activity in memory areas when subjects guessed incorrectly, which suggests that curiosity may enhance memory for surprising new information. This prediction about memory enhancement was confirmed in a behavioral study: Higher curiosity in an initial session was correlated with better recall of surprising answers 1 to 2 weeks later.
4. Wang, Joseph Tao-yi, Michael Spezio and Colin F. Camerer (2010), Pinocchio's Pupil: Using Eyetracking and Pupil Dilation To Understand Truth Telling and Deception in Sender-Receiver Games, American Economic Review, 100(3), 984-1007. (Older versions can be found here, here, here, and here.)
Abstract: We report experiments on sender-receiver games with an incentive for senders to exaggerate. Subjects “overcommunicate”— messages are more informative of the true state than they should be, in equilibrium. Eyetracking shows that senders look at payoffs in a way that is consistent with a level-k model. A combination of sender messages and lookup patterns predicts the true state about twice as often as predicted by equilibrium. Using these measures to infer the state would enable receiver subjects to hypothetically earn 16-21 percent more than they actually do, an economic value of 60 percent of the maximum increment.
5. Wang, Joseph Tao-yi (2010), "Pupil Dilation and Eye-Tracking," in A Handbook of Process Tracing Methods for Decision Research: A Critical Review and User’s Guide, ed. by Michael Schulte-Mecklenbeck, Anton Kuhberger and Rob Ranyard, Psychology Press, 185-204. (book chapter; working paper version)
Abstract: A video-based eye-tracker (e.g., the mobile Eyelink II or Eyelink Remoteof SR Research, Osgoode, Ontario, Canada) uses video cameras to recordthe eye position of human subjects, and hence record pupil dilation and eye movements. The eye-tracker puts video cameras and infrared illuminators in front of the eye to record the position of the eye, cornea reﬂections, and the size of the pupil. Using the movements of one’s pupil (with respect to cornea reﬂections), the eye-tracker tracks the movement of one’s eyes, which is then mapped into locations on the screen by calibration and adjustments for head movements.
6. Ostling, Robert, Joseph Tao-yi Wang, Eileen Chou and Colin F. Camerer (2011), Testing Game Theory in the Field: Swedish LUPI Lottery Games, American Economic Journal: Microeconomics, 3(3), 1-33. (Lead article, Equal contribution with Robert Ostling; online appendix; dataset; working paper version)
Abstract: Game theory is usually difficult to test precisely in the field because predictions typically depend sensitively on features that are not controlled or observed. We conduct a rare such test using field data from the Swedish lowest unique positive integer (LUPI) game. In the LUPI game, players pick positive integers and whoever chose the lowest unique number wins a fixed prize. Theoretical equilibrium predictions are derived assuming Poisson-distributed uncertainty about the number of players, and tested using both field and laboratory data. The field and lab data show similar patterns. Despite various deviations from equilibrium, there is a surprising degree of convergence toward equilibrium. Initial responses can be rationalized by a cognitive hierarchy model.
7. Kuo, Yen and Joseph Tao-yi Wang (2014), Use of Strategy Methods in Experimental Pivotal-Voting Game, Pacific Economic Review, 19(3), 387-400. (working paper version, online appendix)
Abstract: We use the strategy method to conduct laboratory experiments on a nine‐player heterogeneous‐cost voting game. We replicate the underdog and competition effect, but find significantly higher voter turnout rates to be oddddnly partially explained by logit quantal response equilibrium. We examine round‐by‐round changes in cut‐off behaviour and find that voters are highly responsive to historical pivotal events. Voters also respond to past winning and tying, but only as a minority (upsetting the majority), demonstrating an ‘underdog winning effect’, receiving extra utility when winning as a minority. An equilibrium with such asymmetry in utility explains the high minority turnout (and high majority turnout as a best response).
8. Liu, Elaine M., Juanjuan Meng and Joseph Tao-yi Wang (2014), Confucianism and Preferences: Evidence from Lab Experiments in Taiwan and China, Journal of Economic Behavior & Organization, 104, 106-122. (working paper)
Abstract: This paper investigates how Confucianism affects individual decision making in Taiwan and in China. dWe found that Chinese subjects in our experiments became less accepting of Confucian values, such that they became significantly more risk loving, less loss averse, and more impatient after being primed with Confucianism, whereas Taiwanese subjects became significantly less present-based and were inclined to be more trustworthy after being primed by Confucianism. Combining the evidence from the incentivized laboratory experiments and subjective survey measures, we found evidence that Chinese subjects and Taiwanese subjects reacted differently to Confucianism.
9. Mohlin, Erik, Robert Ostling and Joseph Tao-yi Wang (2015), Lowest Unique Bid Auctions with Population Uncertainty, Economics Letters, 134, 53-57. (working paper version)
Abstract: We characterize the unique Poisson–Nash equilibrium of the lowest unique bid auction (LUBA) when the number of bidders is uncertain and follows a Poisson distribution.
10. Lai, Ernest K., Wooyoung Lim and Joseph Tao-yi Wang (2015), An Experimental Analysis of Multidimensional Cheap Talk, Games and Economic Behavior, 91, 114-144. (Online Appendix; working paper version; Former versions titled "Experimental Implementations and Robustness of Fully Revealing Equilibria in Multidimensional Cheap Talk" are here and here.)
Abstract: We design experimental games that capture the logic of Battaglini's (2002) construction of fully revealing equilibrium in multidimensional cheap talk. Two senders transmit information to a receiver over a 2x2 state space. Despite overall misaligned interests, full revelation is achieved in equilibrium by having the senders truthfully reveal along distinct dimensions. Our experimental findings confirm that more information can be extracted with two senders in a multidimensional setting. The extent to which information is transmitted depends on whether dimensional interests are aligned between a sender and the receiver, the sizes of the message spaces, and the specification of out-of-equilibrium beliefs. While inducing interest alignment on the relevant dimensions and restricting the message spaces facilitated equilibrium play and information transmission, having a fully revealing equilibrium that is supported by implausible beliefs reduced the instances in which the equilibrium was played.
11. Hsieh, Fu-Wen and Joseph Tao-yi Wang (2016), "Cheap Talk Games: Comparing Direct and Simplified Replications," Experiments in Organizational Economics (Research in Experimental Economics, Vol. 19), 19-38. (Working paper version)
Abstract: To study strategic information transmission in organizations, we conduct a simplified version (with only three states) of the sender-receiver game experiment designed by Wang, Spezio, and Camerer (2010), in which an informative sender advises an uninformed receiver to take an action (to match the true state), but has incentives to exaggerate. We also have the same subjects play the original five-state game. We find similar “overcommunication” behavior with Taiwanese subjects – messages reveal more information about the true state than what equilibrium predicts – that let us classify subjects into various level-k types. However, results from the simplified version are closer to equilibrium prediction, with more senders robustly classified as level-2.
12. Chen, Chun-Ting, Chen-Ying Huang and Joseph Tao-yi Wang (2018), A Window of Cognition: Eyetracking the Reasoning Process in Spatial Beauty Contest Games, Games and Economic Behavior, 111, 143-158. (Older versions can be found here, here and here.)
Abstract: We study the reasoning process in an environment where final choices are well understood and the associated theory is procedural by introducing two-person beauty contest games played spatially on two-dimensional grid maps. Players choose locations and are rewarded by hitting targets dependent on opponents' choice locations. By tracking subjects' eye movements (lookups), we infer their reasoning process and classify subjects into various levels. More than a half of the subjects' classifications coincides with their classifications using final choices, supporting a literal interpretation of the level-k model for subject's reasoning process. Lookup analyses reveal that the center area is where most subjects initially look at. This sheds light on the level-0 belief. Moreover, learning lookups of a trial on average could increase payoffs of that trial and eliminates roughly 60% of the gap to empirical best response, indicating how valuable lookups can help predict choices.
13. Lin, Po-Hsuan, Andrew Wooders, Joseph Tao-yi Wang and Walter Miao Yuan (2018), Artificial Intelligence, the Missing Piece of Online Education?, IEEE Engineering Management Review, 46(3), 25-28.
Abstract: Despite the recent explosive growth of online education, it still suffers from suboptimal learning efficacy, as evidenced by low student completion rates. This deficiency can be attributed to the lack of facetime between teachers and students, and amongst students themselves. In this article, we use the teaching and learning of economics as a case study to illustrate the application of artificial intelligence (AI) based robotic players to help engage students in online, asynchronous environments, therefore, potentially improving student learning outcomes. In particular, students could learn about competitive markets by joining a market full of automated trading robots who find every chance to arbitrage. Alternatively, students could learn to play against other humans by playing against robotic players trained to mimic human behavior, such as anticipating spiteful rejections to unfair offers in the Ultimatum Game where a proposer offers a particular way to split the pot that the responder can only accept or reject. By training robotic players with past data, possibly coming from different country and regions, students can experience and learn how players in different cultures might make decisions under the same scenario. AI can thus help online educators bridge the last mile, incorporating the benefit of both online and in-person learning.
14. Battaglini, Marco, Ernest K. Lai, Wooyoung Lim and Joseph Tao-yi Wang (2019), The Information Theory of Legislative Committees: An Experimental Analysis, American Political Science Review, 113(1), 55-76. (SOM, working paper version; 1st revision; old version)
Abstract: We experimentally investigate the informational theory of legislative committees (Gilligan and Krehbiel 1989). Two committee members provide policy-relevant information to a legislature under alternative legislative rules. Under the open rule, the legislature is free to make any decision; under the closed rule, the legislature chooses between a member’s proposal and a status quo. We find that even in the presence of biases the committee members improve the legislature’s decision by providing useful information. We obtain evidence for two additional predictions: the outlier principle, according to which more extreme biases reduce the extent of information transmission; and the distributional principle, according to which the open rule is more distributionally efficient than the closed rule. When biases are less extreme, we find that the distributional principle dominates the restrictive-rule principle, according to which the closed rule is more informationally efficient. Overall, our findings provide experimental support for Gilligan and Krehbiel’s informational theory.
15. Wang, Joseph Tao-yi and Wei James Chen (2019), "Investigating Pupil Dilation in Decision Research," Chapter 3 of Handbook of Process Tracing Methods for Decision Research: A Critical Review and User’s Guide, 2nd ed., edited by Michael Schulte-Mecklenbeck, Anton Kuhberger and Joseph G. Johnson, Taylor & Francis, 45-52.
Abstract: This chapter discusses the researchers who are interested in using pupil diameter as a data input in their research but have little or no prior experience. It explores the modern method of measuring pupil diameter, the common environmental control and experimental design that need to be considered, post-experiment data analysis, and the difficulties in interpreting pupil dilation results along with some solutions. The chapter deals with some examples of actual pupil dilation research. Experimental design for pupil dilation research differs from the usual eye-tracking experiments for several reasons. There are some common static measurements that are being reported regarding the trial pupil data: mean pupil dilation, peak pupil dilation, and latency to peak. The main interpretive challenge is to isolate the exact cause of the pupil dilation. In addition to combining measurements, interpretation of pupil dilation can also be aided by relevant theory that explains behavior.
16. Mohlin, Erik, Robert Ostling and Joseph Tao-yi Wang (2020), Learning by Similarity-weighted Imitation in Winnter-takes-all Games , Games and Economic Behavior, 120, 225-245. (SOM, Working paper version; Older versions are here and here)
Abstract: We study a simple model of similarity-based global cumulative imitation in symmetric games with large and ordered strategy sets and a salient winning player. We show that the learning model explains behavior well in both …eld and laboratory data from one such “winner-takes-all” game: the lowest unique positive integer game in which the player that chose the lowest number not chosen by anyone else wins a …xed prize. We corroborate this …nding in three other winner-takes-all games and discuss under what conditions the model may be applicable beyond this class of games. Theoretically, we show that global cumulative imitation without similarity weighting results in a version of the replicator dynamic in winner-takes-all games.
17. Teng, Joshua Chen-Yuan, Joseph Tao-yi Wang and Cheng-Chen Yang (2020), Justice, What Money Can Buy: An Experiment on Primary Social Goods and the Rawlsian Difference Principle, Constitutional Political Economy, 31(1), 45-69. (Working paper version)
Abstract: Many governments and charities adopt Rawlsian difference principle by maximizing the welfare of the least advantaged and giving priority to equality over efficiency. There are two views about which domain the principle should be applied to. The first applies it to the final distribution of income. Previous empirical studies have focused on this but found little evidence supporting it. The other view linked the principle with Rawlsian primary goods: Since the cost of losing primary social goods is huge, people will maximize the benefit of the least advantaged behind the veil of ignorance, such that everyone has access to necessary means. According to the latter reading of Rawls, we experimentally imposed a great cost for losing primary goods, and observed a salient majority of subjects obeying this principle, unlike previous studies finding a minority. Moreover, even if we lowered the cost for losing primary goods, more than one third of the subjects still adopted this principle.
18. Chen, Wei James and Joseph Tao-yi Wang (2020), A Modified Monty Hall Problem, Theory and Decision, 89(2), 151-156. (Working paper version)
Abstract: We conduct a laboratory experiment using the Monty Hall Problem (MHP) to study how simplified examples improve learning behavior and correct irrational choices in probabilistic situations. In particular, we show that after experiencing a simplified version of the MHP (the 100-door version), subjects perform better in the MHP (the 3-door version), compared to the control group who only experienced the 3-door version. Our results suggest that simplified examples strongly induces learning.
19. Lin, Po-Hsuan, Alexander L. Brown, Taisuke Imai, Joseph Tao-yi Wang, Stephanie W. Wang and Colin F. Camerer (2020), General Economic Principles of Bargaining and Trade: Evidence from 2,000 Classroom Experiments, Nature Human Behaviour, 4(9), 917-927. (ViewOnly) (SOM; Data/Code; Working paper version)
Abstract: Standardized classroom experiments provide evidence about how well scientific results reproduce when nearly-identical methods are used. We use a sample of around 20,000 observations to test reproducibility of behavior in trading and ultimatum bargaining. Double auction results are highly reproducible and are close to equilibrium predictions about prices and quantities from economic theory. Our sample also shows robust correlations between individual surplus and trading order, and autocorrelation of successive price changes, which tests different theories of price dynamics. In ultimatum bargaining, the large data set provides sufficient power to identify that equal-split offers are accepted more often and more quickly than slightly unequal offers. Our results imply a general consistency of results across a variety of different countries and cultures in two of the most commonly utilized designs in experimental economics.
20. Camerer, Colin F., Hung-Ni Chen, Po-Hsuan Lin, Gideon Nave, Alec Smith and Joseph Tao-yi Wang (2022), Using Machine Learning to Understand Bargaining Experiments, Bargaining: Current Research and Future Directions, edited by Emin Karagözoğlu and Kyle B. Hyndman, Palgrave Macmillan, Cham, 407-431.
Abstract: We study dynamic unstructured bargaining with deadlines and one-sided private information about the amount available to share (the “pie size”). “Unstructured” means that players can make or withdraw any offers and demands they want at any time. Such paradigms, while lifelike, have been displaced in experimental research by highly structured bargaining because they are hard to analyze. Machine learning comes to the rescue because the players’ unstructured bargaining behavior can be taken as “features” to predict outcomes. Machine learning approaches can accommodate a large number of features and guard against overfitting using test samples and methods such as penalized LASSO regression. In previous research, we found that the LASSO could add power to theoretical variables in predicting whether bargaining ended in disagreement. We replicate this work with higher stakes, subject experience, and special attention to gender differences, demonstrating the robustness of this approach.
21. Fong, Meng-Jhang and Joseph Tao-yi Wang (2023), Extreme (and Non-Extreme) Punishments in Sender-Receiver Games with Judicial Error: An Experimental Investigation, Frontiers in Behavioral Economics, 2, 1096598.
Abstract: In may real world situations, decision-makers have the opportunity to punish informed senders for their biased recommendations, while lie-detection is far from perfect. Hence, we conduct an experiment which incorporates ex post punishment and monitoring uncertainty into the discrete sender-receiver game of Crawford and Sobel (1982), where a knowledgeable sender sends a cheap-talk message to a receiver who determines a policy action. After taking this action, the receiver observes a noisy signal of the true state and can impose a costly punishment on the sender. We vary the strength of punishment from mild (nominal), strong (deterrent) to extreme (potential of losing everything), and vary receiver’s signal uncertainty when punishment is extreme. We find that receivers punish less as the strength of punishment increases, which suggests people care more about wrongly punishing innocent senders harsher than not being able to hand liars harsher punishments they deserve. More importantly, the opportunity of punishment encourages receivers to follow senders more and thus improves overall information transmission and utilization, even though senders need not exaggerate less.
22. Fišar, Miloš, Ben Greiner, Christoph Huber, Elena Katok, Ali Ozkes and the Management Science Reproducibility Collaboration (2023), Reproducibility in Management Science, Management Science, forthcoming. (As Member of the Management Science Reproducibility Collaboration; Online Appendix, OSF)
Abstract: With the help of more than 700 reviewers we
assess the reproducibility of nearly 500 articles published in the journal
before and after the introduction of a new Data and Code Disclosure policy in
2019. When considering only articles for which data accessibility and hard- and
software requirements were not an obstacle for reviewers, the results of more
than 95% of articles under the new disclosure policy could be fully or largely
computationally reproduced. However, for 29% of articles at least part of the
dataset was not accessible to the reviewer. Considering all articles in our
sample reduces the share of reproduced articles to 68%. These figures represent
a significant increase compared to the period before the introduction of the
disclosure policy, where only 12% of articles voluntarily provided replication
materials, out of which 55% could be (largely) reproduced. Substantial
heterogeneity in reproducibility rates across different fields is mainly driven
by differences in dataset accessibility. Other reasons for unsuccessful
reproduction attempts include missing code, unresolvable code errors, weak or
missing documentation, but also soft- and hardware requirements and code
complexity. Our findings highlight the importance of journal code and data
disclosure policies, and suggest potential avenues for enhancing their
1. Chen, Wei (James) and Joseph Tao-yi Wang* (2010), "Epiphany Learning for Bayesian Updating: Overcoming the Generalized Monty Hall Problem" (Corresponding author)
Abstract: We study how people learn the correct action in a probabilistic situation. In particular, we create a modified version of the Monty Hall problem, and conduct laboratory experiments to show how a 100-door variant of the problem helps people learn to play optimally (always switch). Experimental results show that after playing the 100-door variant, subjects obtain an average switching rate above 80% in the 3-door problem, higher than most of the previous work without subject communication and/or competition. Moreover, results from estimating structural learning models using subject-level data show that the individual learning process is more likely to be an epiphany rather than a gradual one.
2. Own, Vivian Tzu-fan and Joseph Tao-yi Wang* (2010), "How Price Tags Affect Willingness-To-Pay---Evidence from the Field (and Lab)"
Abstract: We investigate how posted prices affect consumers' willingness-to-pay (WTP) for real-world products by eliciting the WTP from experienced consumers for water-resisting handbags and consumer electronics accessories before and after people see the price tag. To control for possible experimental artifacts, we elicit WTP with the following procedure: The incentive compatible Becker, DeGroot, Marschak (BDM) mechanism, explanations of the optimal strategy under BDM (truthfully revealing one's valuation), and paid practice rounds with subjects switching roles between buyers and sellers. Though this procedure has successfully minimized the willingness-to-pay and willingness-to accept gap in the literature (which we indeed replicate), we find a moderate but significant increase in WTP for the majority whose initial WTP were lower than the price tag, and a sharp decrease in WTP among the few whose initial WTP were higher than the price tag. This suggests a price effect driven by information regarding potential resale (and repurchase) opportunities. A similar laboratory experiment with college students replicates these findings. This suggests that when firms determine prices or discounts, they might have to care more about the negative “outside-opportunity” effect of low prices instead of the positive “price placebo” effect commonly observed in more controlled environments.
3. Wang, Joseph Tao-yi (2006), "The eBay Market as Sequential Second Price Auctions---Theory and Experiments"
Abstract: We consider the sequential second price auction in which the two highest bids are announced after each stage with unit demand bidders of affiliated private values. We find an efficient symmetric sequential equilibrium in monotonic strategies where bidders bid the expected winning price of the next stage, conditional on being the tied winner in this stage. Such an equilibrium survives in an eBay auction setting, in which bidders are allowed to submit multiple proxy (second-price) bids in a given fixed time period, when all bidders bid only at the last minute, or snipe. We then conduct controlled (laboratory) experiments with the special case of three items sold to bidders with independent private values, using either a sealed-bid second price auction or an eBay auction format. We show the timing of bidding in the eBay auction demonstrate “last minute bidding (sniping)” as reported by Roth and Ockenfels (2002), and early “squatting” coined by Ely and Hossain (2006). Moreover, both formats achieve high efficiency levels and have estimated bid functions seemly as predicted. However, prices decline, but revenue is higher than predicted in the early stages. Individual bids reveal that a significant portion of subjects either bid their valuations or zero in early auctions, even though they bid their valuations in the last one. To explain this, we construct a steps-of-thinking (cognitive hierarchy) model anchored on L0 bidders who always bid their valuations. The combination of various types of bidders explain the data better than that of standard equilibrium theory and shed light on the “declining price anomaly” documented by Ashenfelter (1989) and previously attributed to extreme risk aversion by McAfee and Vincent (1993).
4. Wang, Joseph Tao-yi (2006), "Is Last Minute Bidding Bad?"
Abstract: We demonstrate how last minute bidding on internet auctions is not compatible with a private value setting with a proxy bidding system, but can be rationalized by adding another identical auction. In the repeated eBay auction model, the last minute bidding equilibrium, in which bidders only bid at the last minute in the first auction, is the unique symmetric equilibrium considering monotonic, undominated strategies. Nonetheless, though high valuation bidders efficiently win the items, under strict affiliation, expected revenue is lower in the first auction than in the last, and sellers can increase their revenue by running a joint multi-unit auction.
神經經濟學淺談 (Neuroeconomics: An Introduction) (in Chinese)
實驗經濟學簡介 (What is Experimental Economics?) (in Chinese)
Here are the experimental laboratories I have been involved in:
SSEL - Social Sciences Experimental Laboratory (at Caltech)
Last modified on 十二月 25, 2023