Those performance feedback modes provide useful information for individuals about their achievements relative to others, and may indicate how much they need to improve. Yet, there are other types of performance feedback that seem irrelevant. This is because the feedback does not provide any useful information for individuals seeking to improve their outcomes. Consider the following newsvendor game as an illustrative example. Figure 1 shows the realized demand quantity, the optimal quantity, and the winner’s order quantity per round from a newsvendor game in which 24 players played the game for 30 rounds. The results show that the interim winner’s order quantity and the realized demand quantity are highly correlated; however, they have nothing to do with the (constant) optimal order quantity. The interim winner’s order quantity per round seems an irrelevant input in helping individuals make correct decisions for the long run. For example, if an individual follows the optimal order quantity for all 30 rounds, the average profit per round she would earn is 17.4% higher than she would get by following the interim winner’s order quantity. Besides, as explained above, when information about the realized demand is revealed to the participants, the information about the interim winner’s performance is useless.
Our study is closely related to Song et al.’s (2017), but the benchmark information in our study differs from the best practice information displayed in their study. First, in Song et al. (2017), mid- and bottom-ranked physicians can learn best practices from the top-ranked physicians and utilize those useful experiences and techniques to increase their own productivity. However, the information about the interim winner’s performance in our study represents the best profit and the associated order quantity among all players for a single round only, making it irrelevant benchmark information. Second, in Song et al. (2017), both control and treatment groups have adopted the best practices, and they study the effect of disclosing physicians’ identities on productivity as opposed to no such disclosure, whereas we explore the effect of displaying irrelevant benchmark information without revealing each interim winner’s identity.
Barankay (2012) shows that adding benchmark feedback increases performance on furniture sales compared to only showing rank feedback. The benchmark feedback information he defines is the data on the current performance required to be in the top 10%, 25% and 50%, and benchmark information can indicate feasible outcomes. In our study, the benchmark information refers to the information about the interim winner’s performance only. In Barankay (2012) and other studies mentioned above (e.g., Azmat and Iriberri 2010, Berger and Pope 2011, Charness et al. 2011, Tafkov 2012), results are highly correlated with effort.
Although the irrelevant information, such as benchmark information about the interim winner’s performance per round, may not provide useful and informative guidance for decision makers to make ex-ante decisions, displaying those instances of information might affect managers’ motivation, which could in turn result in improved performance.
2.2 Competition through gamification
Motivation can be generated through gamification, which is defined as using game elements and techniques in non-game contexts (Deterding et al. 2011). In the educational context, gamification has been shown as a powerful tool to create motivation among students regarding enthusiasm and engagement (Kim and Lee 2013). When a game is designed well and fun, it can promote more intrinsic motivation than many other instructional strategies (Lepper and Chabay 1985, Su and Cheng 2015). One key element that can make games fun and motivating is incorporating competition among players (Totty 2005), as competition promotes strong incentives for effort, output, and efficiency without affecting the quality of the output (Cowgill 2015). In a game with competition, the setting itself can make the player feel a sense of achievement (Weinberg and Ragan 1979).
Another popular game element that has been proved to positively affect behavior is providing real or symbolic recognition. For instance, “employee of the month” awards were among the first game structures used by firms (Yates and Wootton 2012) to motivate employees through competition. Leaderboards are another popular tool (Yates and Wootton 2012, Thiebes et al. 2014) and have been shown to improve motivation, engagement, and enjoyment of the learning task (Halan et al. 2010). In a quasi-experimental setting, Markham et al. (2002) show that the introduction of a public recognition program reduces absenteeism by 52%. Moreover, using an experiment conducted online at IBM, Neckermann and Frey (2013) find that the introduction of a hypothetical award significantly affects the stated willingness to contribute to a public good. Parr (2015) shows that rewards, real or symbolic, and acknowledgement both have an impact on attention. Displaying information about an interim winner’s performance can serve as a symbolic award to that interim winner, and it will tend to motivate other players to pay more attention to the game.
2.3 Social comparison theory
One important reason the empirical studies mentioned above can motivate individuals to work harder is social comparison. This theory suggests that individuals are driven to examine their abilities and opinions in comparison to peers (Festinger 1954). Previous research has shown that people engage in social comparison mainly to fulfill certain needs, such as self-evaluation, self-enhancement, and self-improvement (Corcoran at al. 2011, Festinger 1954, Taylor et al. 1996, Wills 1981). For an accurate self-evaluation, individuals tend to use others who are similar as standards. To maintain a positive self-view, individuals are likely to compare themselves with others who are inferior (Wills 1981). Finally, for self-improvement, individuals seem to compare themselves with those who are better in terms of status, capability, and achievements (Shang et al. 2012), because upward social comparison can be motivating (Buunk and Gibbons 2007, Wood 1989). Observing others who perform better tends to make individuals set higher standards, exert greater effort and improve their own performance (Lockwood and Kunda 1997). For example, young children frequently compare themselves with a wide range of other children to learn how to perform tasks (Feldman and Ruble 1977).
Individuals engage in social comparison whenever they encounter information about others (Corcoran et al. 2011). When relative performance feedback is provided, social comparison is likely to be promoted (Gino and Staats 2011). When such feedback is given repeatedly, the impact of social comparisons is likely to be amplified (Gilbert et al. 1995, Mussweiler and Strack 2000). Consequently, individuals improve their performance when they regularly receive relative performance feedback (Gino and Staats 2011). Therefore, we believe that displaying benchmark information should trigger a stronger upward social comparison than occurs when we only show the rank feedback privately, thereby motivating individuals to work harder and improve decision making.
One might be concerned that upward comparison makes people feel that they are inferior to others and has a negative impact on well-being (Suls et al. 2002). This would happen if the feedback information is based on cumulative performance for a number of periods. However, this is not the case when benchmark information is displayed because such information is based on the winner’s performance information for each single round. Therefore, individuals’ self-evaluation does not seem to be affected. Hence, we think the motivating effect of upward social comparison tends to dominate the negative impact.
In conclusion, drawing on the related literature and social comparison theory, we hypothesize that displaying even an irrelevant benchmark information about the interim winner’s performance will lead to better decision making:
Hypothesis 1: Compared to receiving rank feedback alone, individuals will make better decisions on average when they also receive irrelevant benchmark information about the interim winner’s performance publicly.
2.2 Personality traits and decision making
Studies that report average results implicitly assume that individuals are homogeneous. However, individuals are heterogeneous in their characteristics, such as personality, that can potentially modify the effect of a treatment on outcomes. We have suggested that displaying irrelevant benchmark information about the interim winner’s performance tends to motivate individuals to work harder and thereby improve their performance, but this occurs only if those individuals take action toward their goals. In fact, it is possible that some individuals may fail to take effective action. As a consequence, the motivating effect may be smaller for these individuals than those who can better control themselves to take effective actions.