Leader Identification through Networks of Conversational Interruptions

Abstract

Leaders are often identified in empirical studies by either their position in an organizationally defined hierarchy or by survey responses, yet such methods conflate behavioral antecedents and outcomes with behaviors themselves. Furthermore, without an external standard for comparison, it cannot be known to what extent differences in leader assignment or emergence between demographic or other categories are due to behavioral differences or biases in the assessment, selection, training, or rating processes. In this study, we propose the ``interruption network'' as a model of small group structure that is (a) grounded in social status theory, (b) definable in both lab and field groups, (c) minimally impacted by rater bias, and (d) based on assessed behaviors rather than antecedents or outcomes. We show that analysis of interruption networks suggests that the often-reported male bias in leadership attributions is found in the ratings but not in the behavior of the observed individuals: males and females may engage in leadership behaviors equally as often, but males are attributed leadership significantly more. Use of the interruption network to represent small group social status therefore extends well-established research on non-verbal behaviors to an explicitly group-oriented context, supporting theory-based unobtrusive assessment and new ways to address important questions in applied psychology.