Factors Influencing Small-Unit Cohesion in Norwegian Navy Officer Cadets

Abstract

Although cohesion is frequently believed to improve performance in military units, few studies have sought to determine what factors contribute to its formation. In this study, we used a quasi-experimental design to evaluate the effects of several factors on small-unit cohesion in 2 cohorts of Norwegian Navy officer cadets. Cohesion levels in squad-size units were assessed before and after an intensive, week-long sustained operations exercise. Factors found to contribute to increased cohesion levels included the experience of the stressful exercise itself and previous familiarity with other team members. Positive effects on cohesion were also found for personality hardiness and small-unit leadership, suggesting that how stressful experiences get interpreted at the individual and the group levels also influence unit cohesion. Taken together, these findings suggest that cohesion can be increased in small military units by maximizing member familiarity within the units and by providing challenging tasks within an organizational and leadership climate that emphasizes the overall value and positive meaning of such experiences.

Notes

1 Multiple factor analyses of all 20 items with both orthogonal and oblique rotations revealed only one large general factor and found little support for the subdimensions of cohesion (horizontal, vertical, and organizational, further divided along affective and instrumental vectors) posited by Siebold and Kelly (1988). This could reflect cultural differences, or it could be an artifact of the brevity of this measure because it is a short index form of a longer, 79-item measure (Siebold, 1999). At any rate, these factorial results lend further support to the conceptual approach to item selection followed in this study.

2 These same four items were used by Mael and Alderks (1993) to measure unit-level (platoon and squad) cohesion.

3 As many authors have pointed out (e.g., Griffith & Vaitkus, 1999; Gully et al., 1995; Oliver et al., 1999), concerns about level of analysis are important in cohesion studies. For example, if cohesion is a social or group-level phenomenon, perhaps it should be assessed at the group level, as opposed to the recording of individual responses. We agree with this point in principle, especially with regard to the need for appropriate group performance outcome measures. Nevertheless, it remains true that many group-level phenomena can be accurately reflected by individuals reporting on themselves and on fellow group members (including leaders). From this perspective, group cohesiveness as a social phenomenon can be reasonably approximated by the aggregation of member reports about cohesion, as in this study.

4 This model, with the Hardiness × Leadership interaction term, was slightly better in accounting for cohesion variance than was the model with hardiness and leadership alone. Nevertheless, the two models were so similar that the interaction effect, although significant, is probably best regarded as suggestive.