Peer Performance Feedback, Self-Enhancement, and Exploration  

   with Oliver Baumann and Daniel Newark

   (Experiment), Organization Sciene (forthcoming)

We examine how providing decision makers with peer performance information influences their choices between exploration and exploitation over time. Previous work on organization-level learning suggests a high-performing peer would fuel exploration, while a low-performing peer would dampen it. However, underscoring the importance of studying individual-level behavior as building blocks of organization-level behavior, we provide experimental evidence that this basic dynamic is moderated by the extent to which decision makers interpret performance feedback in a self-enhancing way. More specifically, we present two principal findings. First, individuals who receive information about a high-performing peer explore more than those who receive information about a low-performing peer. Second, compared to individuals with a low tendency to self-enhance, individuals with a high tendency to self-enhance are less likely to explore when receiving information about a high-performing peer. In fact, these individuals explore at levels comparable to those who receive information about a low-performing peer. When these individual dynamics are aggregated, our data suggest that an organization that provides peer performance information will experience either the same or less exploration than an organization that does not, with the exact difference depending on the proportion of high self-enhancers within it. These insights into the contingencies and aggregate effects of how individual decision makers interpret and respond to peer performance information are particularly relevant given recent interest in designing less hierarchical organizations that shape employee behavior and innovation through the provision of feedback, rather than through traditional instruments of coordination and control such as incentives or monitoring.

Presented at (excerpt):

• Carnegie School of Organizational Learning Conference 2020
• Academy of Management Annual Meeting 2021
• Research Seminar at ETH Zurich 2021

    Incentives, Attention, and Search

   with Stephan Billinger 

   (Experiment with Eye Tracking), Reject & Resubmit

   In this paper, we examine how different incentives designed to guide individual

   search shape attention and exploration. Incentives that organizations implement to

   guide search towards (or away from) (un)wanted solutions are likely to reduce the

   number of solutions considered, and may thereby hinder exploration. Using an eye-

   tracking experiment, we refine this intuition. We vary incentives that are positive or

   negative and affect the decision-maker or others. Our findings show that incentives

   affecting the decision-maker negatively, or others positively, lead to less exploration.     When these incentives are prevalent, paying more attention to performance   

   feedback is associated with less exploration. This finding proposes an attention-

   based mechanism that explains how different incentives shape exploration. We

   discuss implications for organizations that seek to pursue multiple goals or to

   operate with less-hierarchical structures.

Presented at (excerpt):

• Organization Science Winter Conference 2022
• DRUID Conference 2021
• Academy of Management Annual Meeting 2019
  

   Categories, Generalization, and Exploration

   with Gaël Le Mens and Thomas Woiczyk

   (Model and Experiment), Preparing for submission

   This study examines the role of categories for individual's exploration behavior and

   how category design can be employed to affect the selection of uncertain

   alternatives. In particular, we develop theoretical predictions about how the

   broadness and structure of categories in a given context affect exploration and

   learning about uncertain alternatives based on a computational model. We test

   these predictions in a behavioral study, in which we manipulate the

   categorization of uncertain alternatives. In line with our predictions, we find that   

   fewer, broader categories lead to less exploration of uncertain alternatives and to

   less selection of superior alternatives which is detrimental to performance. The core

   of our mechanism is that categorization of alternatives in distinct categories reduces

   the generalization of negative experiences to other alternatives, whereas

   categorization in fewer, broader categories favors such generalization. Importantly,

   the consequences of inappropriate representation of reality through category

   systems are asymmetric: errors caused by over-specified categories, are likely to be

   corrected over time but errors due to under-specification will persist. We, thus,

   demonstrate the benefits of more distinct categories when designing category

   systems within organizations and discuss important implications for the context of

   fostering innovation within organizations as well as corporate communication more

   generally.

Presented at (excerpt):

• Strategic Management Society Conference 2020
• Academy of Management Annual Meeting 2020

• Society for Judgment and Decision Making Annual Meeting 2020

   Sugarcoating and Exploration in Decentralized Organizational Structures

   with Oliver Baumann and Thorsten Wahle 

   - AOM Best Paper Proceedings, 2022, OMT Division - 

   (Experiment), Preparing for submission

   In this paper, we study how information asymmetries and the opportunity to

   inaccurately report performance feedback in decentralized organizational structures

   influences exploration. In many organizations, exploration is characterized by a two-

   staged process, where a higher-level decision-maker selects a lower-level unit to

   engage in exploration efforts, and a lower-level unit reports back a performance

   outcome. Such decentralized structures are also prone to information asymmetries,

   which are generally seen as undesirable, and aimed to be reduced through

   monitoring. In this study, we provide experimental evidence that information

   asymmetries are not always undesirable for organizations and that monitoring can

   hinder exploration efforts in organizations. The key results from our two experimental

   studies show that, first, lower-level units that take advantage of information

   asymmetries and report performance outcomes inaccurately, explore more and

   experiment more with risky solutions. Second, at the higher level, monitoring, and the

   revelation of inaccurate reporting leads to the avoidance of particular units, less

   exploration and foregone gains of exploration in this domain. This is important as it

   enlarges our understanding on how decentralized structures can foster exploration –

   against the intuition – information asymmetries can be beneficial, and managers

   may rather accept them than trying to reduce them. These insights are particularly

   relevant considering recent interest in ‘flatter’ organizational structures with more

   decision-making authority to lower-level units and less control.

Presented at (excerpt):

• Academy of Management 2022 (accepted)
• Carnegie School of Organizational Learning Conference 2021
• Experiments in Organization Science Seminar 2021