Principled threshold selection for identifying emergent Informational Boundary Spanners (IBSs)

Determine a principled criterion for choosing the threshold on the distribution of informational boundary-spanning Function Executions (FEs) that distinguishes emergent Informational Boundary Spanners (IBSs) from other IBS candidates, such that the threshold is not too low (which would capture occasional performers and reduce sensitivity) and not too high (which would miss significant contributors), thereby ensuring that the identified IBSs meaningfully enhance inter-group information exchange.

Background

The paper introduces a micro-level measurement method for the emergence of Informational Boundary Spanners (IBSs) based on counting each agent’s informational boundary-spanning Function Executions (FEs)—instances where an agent provides new and external information needed by a group.

To classify which IBS candidates truly emerge as effective IBSs, the authors propose applying a threshold on the FE distribution. They note the threshold must be sufficiently high to exclude occasional contributors yet not so high that it excludes significant contributors. However, the authors explicitly state that it is unclear how to select such a threshold, motivating further methodological development.

References

Such threshold or thresholds should be high enough to avoid capturing those IBS candidates that obtain only a few FEs (and as such do not considerably improve inter-group information exchange), but also not too high to avoid not capturing agents that may not have the highest FEs but still contribute significantly to inter-group information exchange. However, it is unclear how to select a threshold that satisfies these conditions.

— Learning to connect in action: Measuring and understanding the emergence of boundary spanners in volatile times  (2405.11998 - Nespeca et al., 2024) in Section 7.1, Experiment 0: Measuring the emergence of informational boundary spanners