Familiarity Breeds Teamwork 28
Running head: FAMILIARITY BREEDS BACKUP
Can familiarity breed backup?
Interactive effects of team efficacy and shared teammate knowledge
Kimberly A. Smith-Jentsch
University of Central Florida
Kurt Kraiger
University of Tulsa
Janis A. Cannon-Bowers
Eduardo Salas
University of Central Florida
Author Note
Kimberly Smith-Jentsch, Department of Psychology; Kurt Kraiger, Department of Psychology; Janis A. Cannon-Bowers, Department of Film and Digital Media and Institute for Simulation and Training; Eduardo Salas, Department of Psychology and Institute for Simulation and Training.
Address correspondence to Kimberly A. Smith-Jentsch,
The views expressed here are those of the authors and do not reflect an official position of the organizations with which they are affiliated.
Abstract
One hundred fifty-two air traffic controllers participated in a cross-sectional study of individual differences in the tendency to request and accept backup. Results indicated that teammate familiarity, supervisory presence, and task experience each accounted for unique variance in peer ratings of on-the-job behavior. Moreover, the relationship between teammate familiarity and behavior appears to have been mediated by perceived team efficacy and shared teammate knowledge. Implications for team development and staffing are discussed.
The last decade has seen an increase in the use of teams in organizations, as well as more focused efforts by researchers to understand the determinants of team effectiveness. One important finding has been that team performance outcomes are determined by more than just the sum of the technical proficiencies held by individual team members. Teamwork processes have been shown to account for a unique amount of variance as well (e.g., Stout, Salas, & Carson, 1994). Backup behavior – shifting workload and helping one’s teammates to perform their roles – has been defined as a core team process in this regard. Effective backup requires that teammates are willing both to provide and to seek assistance (Porter, Hollenbeck, Ilgen, Ellis, West, & Moon, 2003). However, the provision of backup has received much more attention from researchers than has the willingness to accept and to request backup. Moreover, little is known about individual differences that explain when and why certain teammates are more likely to receive backup than others.
Notably, Porter et al. (2003) demonstrated that the personality of those in need of backup played a role in whether or not backup was provided to them by their teammates. For instance, individuals who scored higher on extraversion were provided more backup than those scoring lower on this trait. As an explanation for this finding, the authors argued that extraverts should be more willing to ask for help when they need it. Porter et al. noted that future research should continue to investigate other individual differences beyond personality that may explain which teammates are most likely to receive backup and why. The present study was designed to address this need.
Our approach was similar to that of Porter et al. (2003) in two respects. First, our primary objective was also to better understand the antecedents, rather than the consequences of backup behavior. Second, we too sought to explain individual differences among teammates. However, rather than investigating teammate personality, we studied individual differences in teammate familiarity. Additionally, our dependent variable assessed whether participants tended to accept and to request backup when needed rather than whether, or to whom, they tended to provide backup. Finally, in the present study, behavioral ratings were obtained from teammates in a field setting characterized by high stress, serious consequences of error, and a competitive climate whereby asking for help is seen by some as a weakness; namely, commercial air traffic control (ATC). Like in many team environments, controllers frequently serve as replacement members of teams other than their own primary team. Thus, on any given day, the ATC tower team controlling your departure from an airport may consist of a mixture of familiar and unfamiliar teammates. This presented us with a unique opportunity to study individual differences in the tendency to request and accept backup when needed.
Teammate Familiarity and Backup
As depicted in Figure 1, we proposed that when considering whether or not to ask for or accept backup, individuals consciously or unconsciously weigh the risks of shifting responsibilities to teammates, relative to persisting at their task unassisted. If they feel that the transition of workload can be made quickly, smoothly, and competently, they will be more likely to request and/or to accept backup. In this regard, we suggest that individuals consider their perception of a team’s collective task ability, or “team efficacy,” and the extent to which they agree with their teammates regarding each member’s specific areas of expertise. Specifically, we expected that shared teammate knowledge would strengthen the relationship between team efficacy perceptions and the tendency to request and to accept backup. Furthermore, teammate familiarity was expected to be positively associated with both of these variables.
------
Insert Figure 1 about here
------
Shared mental model theory posits that, as teammates become familiar with one another, they develop a shared understanding of their tasks, roles, equipment, and unique characteristics (e.g., strengths and weaknesses) (Cannon-Bowers, Salas, & Converse, 1993). Shared mental models of teammate characteristics are also a key component of transactive memory theory (Rulke & Rau, 1997; Wegner, 1987). Both theories propose that when teammates share knowledge about one another, they are able to coordinate more efficiently and effectively. In particular, shared teammate knowledge is expected to make social information searches more efficient (Hollingshead, 1998a; 1998b), to allow teammates to anticipate one another’s actions, and to coordinate without the need for overt communication and negotiation (Kleinman & Serfaty, 1989).
However, there has not been consistent support for the notion that teammate familiarity positively affects performance or that it does so through shared cognition (e.g., Goodman & Garber, 1988; Goodman & Leyden, 1991; Mathieu, Heffner, Goodwin, Salas, & Cannon-Bowers, 2000). Restriction of range with respect to the manipulation and measurement of teammate familiarity has been cited as one potential explanation for this. Additionally, it has been noted that global measures of performance effectiveness can mask the behavioral influence of teammate familiarity and shared teammate knowledge.
At least two prior studies have reported findings that suggest, indirectly, that teammate familiarity is associated with backup behavior. First, Shah and Jehn (1993) found that teams of friends were more likely to provide and request assistance than teams of acquaintances. Later, Campion, Papper, and Medsker (1996) found that members of single-identity teams reported being more effective at workload sharing than members of multiple-identity teams. Both friends and members of single-identity teams, by definition, spend more time together than acquaintances or members of multiple-identity teams. Thus, they may have developed greater agreement regarding one another’s relative strengths and weaknesses. However, this is not the only possible explanation for prior findings.
Friends, by definition, differ from acquaintances in that they trust one another and expect to have an ongoing relationship. Members of single-identity teams may or may not trust or care about each other, but they too expect to have a continuing relationship. They may also have a greater stake in the team’s collective success than would replacement members. In the present study, we sought to directly measure familiarity with one’s teammates. Thus, after controlling for team membership, we tested relationships between participants’ tendency to accept and request backup, the frequency with which they worked on that particular team and the length of time they had been acquainted with the members of that team. Furthermore, we investigated the possibility that shared teammate knowledge and perceived team efficacy mediated the familiarity-behavior relationship.
Theoretically, those in greater agreement with their teammates regarding one another’s relative expertise should be able to quickly decide who they believe is both able and willing (since they agree they have the skill to help) to provide backup under a given set of circumstances. Shared teammate knowledge should also reduce the need for lengthy explanations of the backup need and the manner in which backup will be provided. Thus, when knowledge about teammates is shared, shifting tasks during periods of high workload should seem to be less risky in the minds of prospective backup recipients. However, if one believes that the teammates collectively lack sufficient expertise to be helpful, shared teammate knowledge may have the opposite effect. In this regard, we proposed that the influence of shared teammate knowledge on requesting and accepting backup could not be clearly understood without considering perceptions of team efficacy.
Team efficacy has been defined as the belief in a team’s collective capability to perform a specific task (Gully, Incalcaterra, Joshi, & Beaubien, 2002). It is not simply an aggregate of individual teammates’ self-efficacies, but rather a belief in the collective ability of the team as a whole (Bandura, 2000). Although the team efficacy-performance relationship is strongest at the team-level, it has been demonstrated at the individual-level of analysis as well (Gully et al., 2002). Moreover, perceptions of team efficacy can only be aggregated to the team level-of-analysis when there is a high degree of agreement among team members. When team membership is highly variable from day to day, such as it is in the air traffic control environment, teammates may hold very different perceptions of their collective capabilities. Therefore, individual differences in the willingness to request and accept backup should be best predicted by individual-level team efficacy perceptions.
As we have argued earlier, accepting support from another team member involves the risk that shifting responsibility may actually hurt the team’s performance. Thus, individuals should be more receptive to backup when working in a team they believe to be highly competent. However, when task expertise is multi-dimensional and the problems faced by a team are variable and complex, individuals may have a general sense of confidence in their team, yet still be unsure about which of their teammates is best equipped to back them up given a unique type of problem. Moreover, they may be unsure about the manner in which their teammates expect to accomplish the transition of workload. If it is judged that workload cannot be safely shifted without explicit coordination, and time does not permit such coordination, an individual should be less likely to request or accept backup even from teammates they perceive to be highly competent. Thus, for those holding high perceptions of team efficacy, shared teammate knowledge should increase the tendency to accept and request backup. On the other hand, for those holding low perceptions of team efficacy, shared teammate knowledge should decrease the likelihood that backup will be requested. In such a case, individuals will not only doubt their teammates’ ability to provide effective backup, but also whether those teammates (holding similarly low assessments of their ability to help) would agree to provide such assistance if asked. Thus, our first hypothesis stated:
Hypothesis One: Shared teammate knowledge will interact with perceptions of team efficacy to explain one’s tendency to accept and request backup. Specifically, shared teammate knowledge will be positively related to behavior when perceived team efficacy is high and negatively related when team efficacy is low.
As the conceptual model in Figure 1 implies, our second hypothesis tested the notion that team efficacy and shared teammate knowledge mediate the relationship between teammate familiarity and one’s tendency to accept and request backup.
Hypothesis Two: The relationship between teammate familiarity and one’s tendency to accept and request backup will be mediated by the interaction of perceived team efficacy and shared teammate knowledge.
Method
Participants
In the present study, hypotheses were tested using commercial air traffic controllers certified to work in the tower cab. The overall mission of a control tower team is to ensure safe and efficient movement of aircraft on the ground, during take-off and landing, and within a five-mile radius of the airport. The manner in which this team task is partitioned into position responsibilities varies from tower to tower as a function of the volume, type, and complexity of air traffic, as well as the physical layout of the runways and weather conditions. The team task itself largely involves the sequential coordination of individuals performing interdependent and differentiated roles. Full performance level controllers working in the tower cab are fully cross-trained to work each team position and routinely (often daily) rotate through the positions in order to maintain their technical competence. Thus, although each controller in the tower is assigned to a specific position (e.g., ground or local control) at any given point in time, they are certified to work any of the team positions and thus could potentially provide backup to each of their teammates. Despite this fact, controllers are quick to point out the variability in their co-workers’ level of technical knowledge and skill. Moreover, they note that controllers may adopt different strategies or styles of “working traffic” as a result of their relative strengths and weaknesses.
A stratified random sample of air traffic control tower facilities under the jurisdiction of the United States Federal Aviation Administration was selected to participate in the study. All of the facilities sampled reported that they assigned controllers to primary teams. However, they also unanimously stated that controllers frequently served as replacement members of other teams due to absenteeism, scheduled vacations, and special circumstances (unusually heavy traffic or bad weather). Thus, in order to capture the natural variance in teammate familiarity that exists from day to day, we instructed participants to consider the teammates with whom they worked on the day of the study to be their referent when responding to survey items.
Sampling Method and Procedure
Participating facilities were instructed to request all controllers who had worked two randomly selected shifts on the day of the study to complete survey measures during regular working hours. This resulted in an average of 42% (range 40%-45%) of the controllers per facility participating in the study. In total, 207 air traffic controllers from 97 teams working in 51 towers across the United States completed surveys. Facility managers provided tower-level data on team size and supervisory presence. Controllers provided data on their task experience, teammate familiarity, perceptions of team efficacy, and knowledge about their teammates through a computer-based application administered during working hours. In addition, controllers were asked to rate the tendency of each of their focal teammates to accept and request backup when it was legitimately needed based on direct observations while working with them over the previous six months. Participants completed all measures independently. Data were saved to a disk and returned to the researchers by a representative from each facility.