Emergent relationships between team member interpersonal styles and cybersecurity team performance

摘要

A gap in manpower and expertise exists in cybersecurity network defense teams. Little is known about how to compose these defense teams with maximum capability to address a wide variety of hard problems. Team selection can involve a wide array of human factors. We targeted a subset of interpersonal personality traits that in other disciplines, were related to short- and long-term team performance in dyad work teams as well as therapeutic group success. Eight novice cybersecurity teams completed the Interpersonal Adjective Scale-Revised (IAS-R) inventory [31] prior to participating in a 2-day cybersecurity exercise on a simulated network. Teams were rank ordered from 1 to 8 based on their overall team score tracked in a jeopardy-style score board embedded in the exercise. We explored the relationship between team performance and interpersonal style complementarity as well as the degree of interpersonal maladjustment (vector length). First, we tallied the number of complementary interpersonal styles for the top-four and bottom-four performing teams using four theories of complementarity ([7] [9] [26] [27]). No trends emerged to support a possible relationship between the number of complementary interpersonal styles on a given team and respective team performance. Second, we evaluated the mean interpersonal style vector lengths and found that our top performing teams had the highest mean style vector lengths; long style vector lengths is an indicator of potential interpersonal maladjustment. Third, we explored which octant-level traits were predominant in the top- and bottom-performing teams. The top-performing teams tended to have higher counts of extreme expressions of hostile traits and bottom-performing teams tended to have higher counts of moderate expressions of friendly traits. Thus, maladjustment in the hostile quadrants of traits was trending in our top performing teams comparatively. However, small sample sizes and uncontrolled research settings make it difficult to interpret these findings with any certainty. We aimed to provide results using well-established social science theory of interpersonal styles and respective psychometrics for future cybersecurity researchers to build from.