1. Lazer D. et al. 2009. Life in the network: the coming age of computational social science. Science323, 721–723. (doi:10.1126/science.1167742) [PMC free article][PubMed]
2. Vespignani A. 2012. Modelling dynamical processes in complex socio-technical systems. Nat. Phys.8, 32–39. (doi:10.1038/nphys2160)
3. Dyer JR, Johansson A, Helbing D, Couzin ID, Krause J. 2009. Leadership, consensus decision making and collective behaviour in humans. Phil. Trans. R. Soc. B364, 781–789. (doi:10.1098/rstb.2008.0233) [PMC free article][PubMed]
4. Pentland A. 2012. The new science of building great teams. Harv. Bus. Rev.90, 60–69.
5. Choudhury T, Pentland A. 2003. Sensing and modelling human networks using the sociometer. In 2012 16th Int. Symp. on Wearable Computers, p. 216.
6. Barabási AL. 2003. Linked: how everything is connected to everything else and what it means for business, science, and everyday life. New York, NY: Plume.
7. Newman MEJ. 2010. Networks: an introduction. Oxford, UK: Oxford University Press.
8. Wasserman S, Faust K. 1994. Social network analysis: methods and applications. Cambridge, UK: Cambridge University Press.
9. Sparrowe RT, Liden RC, Wayne SJ, Kraimer ML. 2001. Social networks and the performance of individuals and groups. Acad. Manage. J.44, 316–325. (doi:10.2307/3069458)
10. Girvan M, Newman MEJ. 2002. Community structure in social and biological networks. Proc. Natl Acad. Sci. USA99, 7821–7826. (doi:10.1073/pnas.122653799) [PMC free article][PubMed]
11. Watts DJ, Dodds PS, Newman MEJ. 2002. Identity and search in social networks. Science296, 1302–1305. (doi:10.1126/science.1070120) [PubMed]
12. Newman MEJ, Park J. 2003. Why social networks are different from other types of networks. Phys. Rev. E.68, 036122 (doi:10.1103/PhysRevE.68.036122) [PubMed]
13. Bird C, Pattison D, D'souza R, Filkov V, Devanbu P. 2008. Latent social structure in open source projects. In Proc. of the 16th ACM SIGSOFT Int. Symp. on Foundations of Software Engineering. SIGSOFT '08/FSE-16, pp. 24–35. New York, NY: ACM.
14. Ahn YY, Bagrow JP, Lehmann S. 2010. Link communities reveal multiscale complexity in networks. Nature466, 761–764. (doi:10.1038/nature09182) [PubMed]
15. Mucha PJ, Richardson T, Macon K, Porter MA, Onnela JP. 2010. Community structure in time-dependent, multiscale, and multiplex networks. Science328, 876–878. (doi:10.1126/science.1184819) [PubMed]
16. Palla G, Barabási AL, Vicsek T. 2007. Quantifying social group evolution. Nature446, 664–667. (doi:10.1038/nature05670) [PubMed]
17. Brooks FP., Jr 1995. The mythical man-month. Boston, MA: Pearson Education.
18. Faraj S, Sproull L. 2000. Coordinating expertise in software development teams. Manage. Sci.46, 1554–1568. (doi:10.1287/mnsc.46.12.1554.12072)
19. Rising L, Janoff NS. 2000. The Scrum software development process for small teams. IEEE Softw.17, 26–32. (doi:10.1109/52.854065)
20. Milojević S. 2014. Principles of scientific research team formation and evolution. Proc. Natl Acad. Sci. USA111, 3984–3989. (doi:10.1073/pnas.1309723111) [PMC free article][PubMed]
21. de Montjoye YA, Stopczynski A, Shmueli E, Pentland A, Lehmann S. 2014. The strength of the strongest ties in collaborative problem solving. Sci. Rep.4, 5277 (doi:10.1038/srep05277) [PMC free article][PubMed]
22. Clauset A, Young M, Gleditsch KS. 2007. On the frequency of severe terrorist events. J. Confl. Resolut.51, 58–87. (doi:10.1177/0022002706296157)
23. Bohorquez JC, Gourley S, Dixon AR, Spagat M, Johnson NF. 2009. Common ecology quantifies human insurgency. Nature462, 911–914. (doi:10.1038/nature08631) [PubMed]
24. Henrich J. 2004. Demography and cultural evolution: how adaptive cultural processes can produce maladaptive losses: the Tasmanian case. Am. Antiquity69, 197–214. (doi:10.2307/4128416)
25. Derex M, Beugin MP, Godelle B, Raymond M. 2013. Experimental evidence for the influence of group size on cultural complexity. Nature503, 389–391. (doi:10.1038/nature12774) [PubMed]
26. Andersson C, Read D. 2014. Group size and cultural complexity. Nature511, E1 (doi:10.1038/nature13411) [PubMed]
27. Derex M, Beugin MP, Godelle B, Raymond M. 2014. Derex et al. reply. Nature511, E2 (doi:10.1038/nature13412) [PubMed]
28. Price DdS. 1963. Little science, big science. New York, NY: Columbia University Press.
29. Pao ML. 1992. Global and local collaborators: a study of scientific collaboration. Inf. Process Manage.28, 99–109. (doi:10.1016/0306-4573(92)90096-I)
30. Hudson J. 1996. Trends in multi-authored papers in economics. J. Econ. Perspect.10, 153–158. (doi:10.1257/jep.10.3.153)
31. Latane B, Williams K, Harkins S. 1979. Many hands make light the work: the causes and consequences of social loafing. J. Pers. Social Psychol.37, 822–832. (doi:10.1037/0022-35184.108.40.2062)
32. Harkins SG. 1987. Social loafing and social facilitation. J. Exp. Soc. Psychol.23, 1–18. (doi:10.1016/0022-1031(87)90022-9)
33. Karau SJ, Williams KD. 1993. Social loafing: a meta-analytic review and theoretical integration. J. Pers. Soc. Psychol.65, 681–706. (doi:10.1037/0022-35220.127.116.111)
34. Denis JL, Lamothe L, Langley A. 2001. The dynamics of collective leadership and strategic change in pluralistic organizations. Acad. Manage. J.44, 809–837. (doi:10.2307/3069417)
35. Johnstone RA, Manica A. 2011. Evolution of personality differences in leadership. Proc. Natl Acad. Sci. USA108, 8373–8378. (doi:10.1073/pnas.1102191108) [PMC free article][PubMed]
36. Contractor NS, DeChurch LA, Carson J, Carter DR, Keegan B. 2012. The topology of collective leadership. Lead. Q.23, 994–1011. (doi:10.1016/j.leaqua.2012.10.010)
37. Katzenbach JR, Smith DK. 1993. The wisdom of teams: creating the high-performance organization. Boston, MA: Harvard Business Press.
38. Delmar F, Shane S. 2006. Does experience matter? The effect of founding team experience on the survival and sales of newly founded ventures. Strateg. Organ.4, 215–247. (doi:10.1177/1476127006066596)
39. Lindbeck A, Snower DJ. 2000. Multitask learning and the reorganization of work: from Tayloristic to holistic organization. J. Lab. Econ.18, 353–376. (doi:10.1086/209962)
40. Postrel S. 2002. Islands of shared knowledge: specialization and mutual understanding in problem-solving teams. Organ. Sci.13, 303–320. (doi:10.1287/orsc.13.3.303.2773)
41. Crowston K, Wei K, Howison J, Wiggins A. 2012. Free/Libre open-source software development: what we know and what we do not know. ACM Comput. Surv.44, 1–35. (doi:10.1145/2089125.2089127)
42. Scholtes I, Mavrodiev P, Schweitzer F. In press From Aristotle to Ringelmann: a large-scale analysis of team productivity and coordination in Open Source Software projects. Empir. Softw. Eng. (doi:10.1007/s10664-015-9406-4)
43. Albrecht AJ, Gaffney JE Jr. 1983. Software function, source lines of code, and development effort prediction: a software science validation. IEEE Trans. Softw. Eng.9, 639–648. (doi:10.1109/TSE.1983.235271)
44. Rosenberg J. 1997. Some misconceptions about lines of code. In Proc. of Fourth Int. Symp. on Software Metrics, pp. 137–142. New York, NY: IEEE.
45. Koch S, Schneider G. 2002. Effort, co-operation and co-ordination in an open source software project: GNOME. Inf. Syst. J.12, 27–42. (doi:10.1046/j.1365-2575.2002.00110.x)
46. Alali A, Kagdi H, Maletic J. 2008. What's a typical commit? A characterization of open source software repositories. In Proc. of the 16th IEEE Int. Conf. on Program Comprehension, 2008, pp. 182–191. New York, NY: IEEE.
47. Crowston K, Howison J, Annabi H. 2006. Information systems success in free and open source software development: theory and measures. Softw. Process: Improv. Pract.11, 123–148. (doi:10.1002/spip.259)
48. Subramaniam C, Sen R, Nelson ML. 2009. Determinants of open source software project success: a longitudinal study. Decis. Support Syst.46, 576–585. (doi:10.1016/j.dss.2008.10.005)
49. Ghapanchi AH, Aurum A, Low G. 2011. A taxonomy for measuring the success of open source software projects. First Monday16 (doi:10.5210/fm.v16i8.3558)
50. Crowston K, Howison J. 2005. The social structure of free and open source software development. First Monday10 (doi:10.5210/fm.v10i2.1207)
51. Mockus A, Fielding RT, Herbsleb JD. 2002. Two case studies of open source software development: apache and mozilla. ACM Trans. Softw. Eng. Methodol.11, 309–346. (doi:10.1145/567793.567795)
52. Wang D, Song C, Barabási AL. 2013. Quantifying long-term scientific impact. Science342, 127–132. (doi:10.1126/science.1237825) [PubMed]
53. Kalliamvakou E, Gousios G, Blincoe K, Singer L, German DM, Damian D. 2014. The promises and perils of mining GitHub. In Proc. of the 11th Working Conf. on Mining Software Repositories, pp. 92–101. New York, NY: ACM.
54. Horwitz SK, Horwitz IB. 2007. The effects of team diversity on team outcomes: a meta-analytic review of team demography. J. Manag.33, 987–1015. (doi:10.1177/0149206307308587)
The 2016 Football (Soccer) European Championships were notable for the emergence and success of two smaller countries (i.e., Iceland and Wales) competing among the giants of the sport. Commentaries about their accomplishments quite often focused on the teams’ ability to work together as a cohesive unit to overcome any deficiencies in individual talent. For example, in a preview of the Icelandic team leading up to the tournament, the magazine WorldSoccer noted:
Since Lars Lagerback took over as coach in October 2011 he has stuck to the 4-4-2 system that he favoured for so many years with his native Sweden. With his current team, the emphasis has very much been on cohesion and team spirit, both in defence and attack … . The old saying that “a chain is never stronger than its weakest link” is acknowledged by everybody on the team. They all accept that each of them has to give 100 per cent, every game; 95 per cent for a side like Iceland is not enough on the big stage. No one is too big for the team.
(Hallgrimsson, 2016, paras. 13 and 15)
The importance of group cohesion is shared among many performance contexts including sport (e.g., Eys, Loughead, Bray, & Carron, 2009), business (e.g., Tekleab, Karaca, Quigley, & Tsang, 2016), military (e.g., Kanesarajah, Waller, Zheng, & Dobson, 2016), and music (e.g., Dobson & Gaunt, 2015). As a result, researchers and practitioners working with performance groups also attempt to facilitate perceptions of cohesion through the process of team building. This article focuses on the physical activity context and will provide an overview of the definition and conceptualization of cohesion, identify measurement tools used to assess athletes’ perceptions of cohesion in sport, highlight the extant literature supporting the importance of cohesion in this context, and discuss the suggestions and protocols that are considered to build high functioning teams within the sport environment (i.e., team building).
Definition and Conceptual Model of Cohesion
Generally speaking, cohesion represents the strength of the bonds among group members or, more informally, the degree to which individuals stick together (Carron & Eys, 2012). This group property has been the subject of considerable research over the past 60 years and definitions have indicated differing approaches to understanding cohesion. For example, Gross and Martin (1952) suggested that cohesion represents the collective resistance to disruption of the group (i.e., the degree to which the group can withstand outside pressures or unfavorable events). Alternatively, Festinger, Schachter, and Back (1963) defined cohesion as the sum of all the forces that cause members to be attracted to, and remain in, the group, and also considered these forces to be related to task and social aspects of the environment.
In sport and exercise research, the most accepted definition of cohesion was provided by Carron, Brawley, and Widmeyer (1998): “a dynamic process which is reflected in the tendency for a group to stick together and remain united in the pursuit of its instrumental objectives and/or for the satisfaction of member affective needs” (p. 213). This definition implies several characteristics of cohesion that include an ability to change over the span of group development (i.e., dynamic), a focus on both task (i.e., instrumental objectives) and social aspects of the group (i.e., member affective needs), and, relatedly, an assumption that it is multidimensional.
With respect to the latter points, and following from the varied approaches of earlier cohesion research, Carron, Widmeyer, and Brawley (1985) proposed a four dimension conceptual model that encompasses two different perceptual orientations (i.e., individuals’ perceptions of their own attractions to the group as well their perceptions about the degree to which the group is integrated) regarding two broad aspects of the group environment (i.e., task and social concerns). In combination, the four dimensions represent individuals’ perceptions of their (a) attractions to task aspects of the group (ATG-T), (b) attractions to social aspects of the group (ATG-S), (c) group’s integration regarding task objectives (GI-T), and (d) group’s integration regarding social objectives (GI-S).
Another interesting aspect regarding the concept of cohesion relates to the dynamism of individuals’ perceptions of their group. McEwan and Beauchamp (2014) proposed that cohesion is an emergent state resulting from (and influencing) other behavioral processes in which the team engages (e.g., teamwork processes). In this sense, cohesion is proposed to be an outcome/antecedent of several group processes (as opposed to being a process unto itself). Regardless, it is interesting to consider whether the various dimensions of cohesion differ with respect to the speed and/or level with which they initially emerge within a group and their ongoing stability. There is some support in the extant literature to suggest that all dimensions of group cohesion do not progress in lockstep. Arrow, Poole, Henry, Wheelen, and Moreland (2004) proposed that group members’ attractions to their group have elements that develop at different speeds. More global attractions to the group are proposed to develop quickly while more specific interpersonal attractions (i.e., among group members) need more time to be fostered.
In a physical activity context, Dunlop, Falk, and Beauchamp (2012) tracked 46 group exercise classes and assessed participants’ perceptions of the four dimensions of cohesion during the 2nd, 5th, and 8th week of the session. They found that perceptions of task cohesion remained relatively stable across exercise sessions, while social cohesion perceptions were more variable over those time points. The researchers suggested that their results had implications toward group interventions in exercise (i.e., opportunities to facilitate social connections within the classes) and provided support that cohesion perceptions are malleable. This result (i.e., greater stability for task cohesion perceptions vs. social cohesion) is consistent with Leeson and Fletcher (2005), who examined cohesion perceptions of 219 elite female netball players across four time points in a competitive season.
Measurement of Cohesion in Sport
The body of knowledge pertaining to cohesion in sport has been aided by several attempts to measure athletes’ perceptions of this group property. These attempts include the Sport Cohesiveness Questionnaire (Martens, Landers, & Loy, 1972), the Multidimensional Sport Cohesion Inventory (Yukelson, Weinberg, & Jackson, 1984), and the Group Environment Questionnaire (Carron, Brawley, & Widmeyer, 2002; Carron et al., 1985). The Group Environment Questionnaire (GEQ) has received the most attention and is the operationalization of the four dimensions of cohesion outlined in the previous section. Specifically, the GEQ is an 18-item measure assessing athletes’ perceptions of their attractions to social (5 items) and task (4 items) aspects of the group, as well as their perceptions of how integrated their group is from both social (4 items) and task (5 items) perspectives. Over time, evidence has been provided regarding the validity and reliability of responses to this assessment tool (see Carron et al., 1998; Carron et al., 2002, for summaries), though certain limitations have been identified. For example, Eys, Carron, Bray, and Brawley (2007) noted that the strategy of using both positively and negatively worded items might create problems for the internal consistency of certain dimensions.
Furthermore, as Carron et al. (2002) noted, “The GEQ was specifically developed, its psychometric properties investigated, and norms established with recreational and competitive sport teams composed of North American female and male athletes between the ages of approximately 18 to 30 years” (p. 39) and encouraged careful consideration of the context specificity of the questionnaire. To this end, researchers have translated and adapted the GEQ to ensure they had a relevant measure of cohesion for their population. As just a few examples, Heuzé and Fontayne (2002) used the GEQ as the basis for a French language cohesion questionnaire (Questionnaire sur l’Ambiance du Group), while Estabrooks and Carron (2000) adapted the measure for use in an exercise class context (Physical Activity Group Environment Questionnaire).
More recently, efforts have been made to examine cohesion in younger athletes including youth (approximately 12 to 17 years of age; Youth Sport Environment Questionnaire; Eys et al., 2009) and children (approximately 9 to 12 years of age; Child Sport Cohesion Questionnaire; Martin, Carron, Eys, & Loughead, 2012). Eys and colleagues (2009) noted several advantages of developing age-appropriate cohesion assessment tools including increased readability. Furthermore, for both questionnaires, the researchers found evidence that younger populations did not distinguish between group integration perceptions and their attractions to the group, but rather viewed their group more globally with respect to task and social cohesion (two dimensions vs. four dimensions). Overall, the efforts of researchers to develop appropriate measures of cohesion have led to a large body of literature within sport. The following section briefly highlights this information.
Research on Cohesion in Sport
Without question, cohesion has been the most heavily researched group dynamics concept in sport psychology. The research questions have tackled a variety of issues including the relationship of cohesion with individual cognition/affect/behavior (e.g., individual effort), other features of the group environment (e.g., motivational climate) and structure (e.g., leadership, roles), and performance. In the following sections, examples of this research are provided to highlight the importance of this emergent state, though we note more extensive coverage can be found in other texts (e.g., Carron & Eys, 2012).
Cohesion and the Individual Athlete
Research linking perceptions of cohesion to important individual correlates has been extensive and includes cognitive, affective, and behavioral variables. For example, from a cognitive perspective, Shapcott and Carron (2010) found that the attributions athletes make regarding team performance were related to task cohesion. As one specific aspect, athletes who had higher perceptions of task cohesion attributed team failures to causes that were controllable and changeable (a more positive attributional approach). Bruner, Eys, Wilson, and Côté (2014) undertook a study to examine cohesion as it relates to positive youth development. Their findings positively linked both task and social cohesion to the development of personal and social skills, initiative, cognitive skills, and goal setting practices.
Affective variables have also been considered and, for the most part, the links have been beneficial. Several studies have examined the association between team cohesion and individual athlete satisfaction. Illustrative of this relationship, Spink, Nickel, Wilson, and Odnokon (2005) found that athletes’ perceptions of how integrated their team was regarding task aspects (GI-T) were positively related to their satisfaction with the group’s contributions and coordination. More recently, Wolf, Eys, and Kleinert (2015) found that greater cohesion was predictive of athletes’ facilitative interpretations of their precompetitive state anxiety symptoms over and above the contributions of other important variables (e.g., trait anxiety).
Finally, the perceptions individuals hold regarding the cohesion of their team are believed to influence their behaviors. Earlier research provides evidence of positive links with key variables in the sport environment, including adherence (e.g., returning the following season to one’s team; Spink, Wilson, & Odnokon, 2010), sacrifice behaviors (e.g., putting aside personal goals for team goals; Prapavessis & Carron, 1997), and social loafing (McKnight, Williams, & Widmeyer, 1991). Furthermore, Bruner, Boardley, and Côté (2014) found that perceptions of task and social cohesion played differential mediating roles between social identity and both pro- and anti-social behaviors toward teammates and opponents. For example, Bruner and colleagues found that stronger perceptions of social identity expressed by athletes were positively related to task cohesion that, in turn, were related to greater prosocial behaviors and lesser antisocial behaviors toward teammates. In contrast, social cohesion perceptions promoted by stronger social identities were predictive of more antisocial behaviors toward opponents.
Cohesion and the Team Environment
Given that cohesion is an emergent group state, it is not surprising that researchers have examined it in light of other important group variables. The volume of studies is too large to cover in-depth within this article, but the existing literature highlights numerous associations with structural, leadership, and environmental variables. From a structural standpoint, greater cohesion has been positively linked with perceptions of group status and roles. For example, Jacob and Carron (1998) found that athletes perceiving higher cohesion attached less importance to status differences within their team. From a different vantage point, participants in a rugged wilderness trek perceived greater cohesion when group members had congruent perceptions of the status structure in their group (Eys, Ritchie, Little, Slade, & Oddson, 2008). With respect to roles, Carron and Eys (2012) summarized that cohesion and role perceptions (e.g., role ambiguity, acceptance, and performance) appear to act on each other in a reciprocal fashion, though Bosselut, McLaren, Eys, and Heuzé (2012) found that youth athletes’ perceptions of social cohesion were predictive of their subsequent perceptions of role ambiguity.
Leaders play an essential role in the emergence of cohesion within the group. The degree to which leaders (both coaches and athlete/peer leaders) demonstrate behaviors related to training and instruction, social support, and the provision of positive feedback, as well as engage their followers via a democratic style (vs. autocratic), is positively related to perceptions of cohesion (Jowett & Chaundy, 2004; Vincer & Loughead, 2010). In addition, coaches not only have responsibilities for interacting effectively with each individual athlete, they also need to act in a manner that is helpful in creating a positive motivational climate. Coaches who provide for a stronger task-involving motivational climate have athletes who perceive greater task and social cohesion (Eys, Jewitt, Evans, Wolf, Bruner, & Loughead, 2013; Horn, Byrd, Martin, & Young, 2012). In contrast, less cohesion is perceived when an ego-involving climate is promoted. On the basis of these previous findings, McLaren, Eys, and Murray (2015) conducted an intervention with youth soccer coaches to educate them about what constitutes a positive motivational climate and to provide strategies for them to use throughout the season. Compared to a control group, athletes whose coaches took part in the intervention perceived a stronger task-involving motivational climate as well as greater perceptions of cohesion by the end of the season.
Cohesion and Performance
The question pertaining to whether cohesion is linked to team performance has stretched as far back as the 1960s, with individual sets of empirical results yielding a somewhat ambiguous picture of this relationship. In an attempt to rectify this situation, Carron, Colman, Wheeler, and Stevens (2002) conducted a meta-analysis of sport studies to determine the general relationship between cohesion and performance as well as potential moderators of this relationship. Specifically, Carron and colleagues examined whether the cohesion-performance relationship differed with respect to type of cohesion (task vs. social cohesion), type of sport (interdependent vs. individual sports), gender (male vs. female), skill level and age, and the direction of the relationship using any lagged longitudinal datasets that were available (cohesion leading to performance vs. performance leading to cohesion). Overall, the researchers found that there was a moderate, positive, and significant relationship between cohesion and performance (effect size = 0.655). This particular relationship held regardless of type of cohesion/sport, skill level, or direction of the relationship. However, there was a moderating effect of gender. In essence, while still significant for males (effect size = 0.556), the positive relationship between cohesion and performance was stronger for females (effect size = 0.949). A follow-up meta-analysis (Filho, Dobersek, Gershgoren, Becker, & Tenenbaum, 2014), examining studies conducted between 2000 and 2010, further supported the general positive relationship between these two variables as well as the moderating effect of gender. However, Filho and colleagues demonstrated there were some differences in the strength of the relationship based on skill level and sport type.
The finding that gender moderates the cohesion-performance relationship was discussed by the groups of researchers. Carron and Colleagues (2002) suggested that this might be important practical knowledge for coaches and sport psychology professionals to consider when working with teams. From a research perspective, Filho and Colleagues (2014) encouraged investigators to “focus on asking ‘why’ (e.g., Why do women and men differ in cohesion dynamics?) to provide explanation of the mediating mechanisms underlying gender idiosyncrasies” (p. 174). This question pertaining to why there may be gender differences was pursued in a qualitative study conducted by Eys and Colleagues (2015). These researchers interviewed 22 Canadian and German coaches who had experience coaching both male and female competitive sport teams over the course of their careers. The researchers asked coaches to comment on the findings and to offer their perspectives regarding why cohesion may be a more important group property for female teams as compared to males. While it is beyond the scope of this article to highlight the results in their totality, coaches tended to agree with the empirical results in the sense that they believed that cohesion was important for both males and females, but that there is a tendency for it to be more important in female teams. Furthermore, coaches offered interesting ideas that could form the basis for future research questions. For example, some coaches observed that the direction of the cohesion-performance relationship might differ for males and females; specifically, that cohesion may drive performance for females while performance may drive perceptions of cohesion for males. This is an interesting proposition that has not yet been tested in the previous meta-analyses. As another example, coaches also felt that there may be temporal differences in the development of cohesion. In essence, male and female teams may differ with respect to the speed that cohesion is facilitated (e.g., faster to develop in male teams).
There are a few limitations to previous research examining the cohesion-performance relationship included in the previous meta-analyses. These include a reliance on young adult populations (+18 years), cross-sectional designs, and sub-elite competitive levels. Benson, Šiška, Eys, Priklerovád, and Slepičkab (2016) sought to address some of these issues in a prospective investigation of the cohesion-performance relationship with elite Czech and Slovak Republic youth football (soccer) and handball teams. Their study included 246 athletes from 18 teams whose perceptions of cohesion were obtained at mid-season and late season along with their team’s performance. In contrast to the general tone of the extant literature suggesting that cohesion leads to performance, Benson and colleagues found evidence that performance outcomes drive perceptions of cohesion in elite youth sport teams. This finding opens up several research questions regarding this relationship across sport and the researchers encouraged continued investigation of the psychological mechanisms (i.e., mediators) and boundary conditions (i.e., moderators) of the cohesion-performance relationship. Certainly, their study had several limitations (e.g., Czech and Slovak Republic athletes only, predominantly male, limited number of sports). Regardless, their result suggesting that performance leads to cohesion in the elite youth sport environment is tantalizing within a body of research that often suggests a bi-directional relationship and/or promotes cohesion as a performance enhancing necessity.
Cohesion as a Potential Disadvantage
As noted, cohesion is believed to be a force for the good of the group. As previous sections have highlighted, cohesion is associated with several important personal, team, and leadership factors, as well as team performance. However, several researchers have cautioned that there are negative aspects to cohesion that need to be considered. This is a concern that is shared and identified by athletes as well. For example, Hardy, Eys, and Carron (2005) asked 105 intercollegiate athletes if they viewed any downsides to group cohesion and, if so, to further discuss the specific issues. Overall, 56% of the athletes queried noted that they saw potential disadvantages to high social cohesion, and 31% indicated potential problems to high task cohesion. It is important to note that many of the issues raised by the athletes appeared to be interpreted in light of an imbalance of team cohesion (i.e., high social cohesion with a relatively lower amount of task cohesion and vice versa). However, the perceived disadvantages of high social cohesion included the potential for communication problems among friends (e.g., afraid to be critical of those you are close with), challenges in fully focusing on the task at hand (e.g., social issues dominating task concerns), and the exclusion of those individuals who do not adhere to the social norms of the group. From a task perspective, the challenges included perceived increases in pressure to perform as well as decreased social and personal enjoyment.
Rovio, Eskola, Kozub, Duda, and Lintunen (2009) added further support to the suggestion that group cohesion can be problematic at times. They conducted a qualitative study with an ice-hockey team over the course of one competitive season. In this case study, the team’s performance decreased as the season progressed though group cohesion appeared to be rather resilient. They suggested that the high social cohesion present on the team might have posed some challenges. In particular, they highlighted the occurrence of several established group dynamics phenomena (i.e., pressures to conform, group polarization, groupthink) that may have led to lower standards of performance. Overall, the issues raised in the Hardy et al. (2005) and the Rovio et al. (2009) studies are in line with those raised in a review by Pescosolido and Saavedra (2012), indicating that cohesion can be yet another strong force to contend with in the group environment that can lead to a reluctance on the part of the individual to violate strong normative pressures to be a team player.
Cohesion as a Target for Intervention
Although there are specific instances in which too much cohesion could be detrimental to a team, the overwhelming evidence suggests that strong group cohesion (both task and social) is a desirable emergent state. The previous section outlining research on cohesion in sport highlighted the many positive individual and group correlates, with arguably the most important connection being the positive association between cohesion and performance. As a result, there have been many attempts designed to enhance athletes’ perceptions of cohesion in their sport teams. In particular, these attempts to improve team effectiveness (i.e., task cohesion) and enhance interpersonal relationships (i.e., social cohesion) are referred to as team building. Brawley and Paskevitch (1997) provided a specific definition of team building for physical activity contexts that described it as a “method of helping the group to (a) increase effectiveness, (b) satisfy the needs of its members, or (c) improve work conditions” (pp. 13–14).
There is evidence to suggest that cohesion has been the primary target (vs. other group constructs) for team-building interventions. In a unique study examining the origins of team building in sport, Bruner, Eys, Beauchamp, and Côté (2013) used citation network and citation path analyses to determine the influential texts and articles that have driven team-building research. Essentially, citation network analysis determines the interconnectedness of citations among a series of publications and determines the most central or influential texts. In parallel, citation path analysis links key texts over time to provide a picture of the evolution of thinking around a particular topic. As Bruner and Colleagues (2013) noted, these two analyses “hold considerable promise to enhance an understanding of [team building] in sport by identifying bodies of literature, and trends, that have shaped the field as well as identifying potential restrictions or omissions that have emerged as the field of enquiry developed” (pp. 31–32).
The major finding from Bruner and Colleagues (2013) was that the extant literature on team building in sport is largely driven by cohesion-focused research. In particular, the work conducted by Carron and colleagues (e.g., Brawley, Carron, & Widmeyer, 1987; Carron et al., 1985) was predominant throughout both the citation network and path analyses. On one hand, this result suggests the importance of cohesion as an outcome of team building. On the other hand, the result supports Bruner et al.’s (2013) cautionary statement that perhaps the field of sport psychology is too narrow with respect to its approach to team building, both in terms of topic (i.e., cohesion) and use of the extant literature (i.e., not giving due consideration to other fields such as organizational psychology). This point was reiterated by McEwan and Beauchamp (2014) in their review of teamwork processes in sport. They noted that team-building processes should move beyond solely considering cohesion and target additional teamwork behaviors such as coordination, cooperation, and communication.
Critiques concerning the narrowly focused nature of team-building processes aside, if these protocols are focused predominantly on cohesion, what can we say about the effectiveness of intervening with sport teams? Martin, Carron, and Burke (2009) conducted a meta-analysis to answer this very question. Their analysis included 17 studies and 180 effect sizes emanating from the data in these investigations. Martin and colleagues found a moderate positive effect for team-building interventions when taken in totality across several dependent variables (e.g., social cohesion, task cohesion, performance, enhanced cognitions, roles, anxiety). However, follow-up moderation tests yielded several interesting findings. First, the researchers found that interventions using team goal setting had larger effects than those interventions that took a broader approach to team-building activities (i.e., targeting several components). They surmised that more positive results might be the product of fewer activities that athletes can truly focus on. Second, consistent with past research in both sport psychology and organizational psychology, interventions were less effective when they were shorter in length (i.e., less than two weeks). Third, Martin et al. (2009) found that team building was particularly effective with independent sport teams (vs. interactive sport teams), but noted that this effect may be due to greater room for developing group interaction in sports that may traditionally offer less opportunities (i.e., a ceiling effect for the interactive teams). Finally, the impact of team-building interventions on perceptions of cohesion (both task and social) were rather muted, which the researchers found interesting given that practitioners and researchers often use team building with the hopes of increasing group cohesion. Certainly, this is a finding with implications that require future research to disentangle and consider regarding the process and targeted outcomes of team building. In the following section, information pertaining to the team-building protocols used in sport are described in further detail.
Team Building Protocols in Sport and Exercise
Given the complex nature of group dynamics and team development, a wide variety of factors must be considered during the creation and delivery of team-building protocols. Thus, researchers and practitioners have taken numerous approaches that have varied on their conceptual basis, delivery medium, types of activities, and outcome measures. This section provides an overview of such protocols and approaches that have been undertaken in the team-building literature in sport and exercise.
At a general level, team-building protocols have largely been categorized as either direct or indirect, based on the role that the interventionist plays in the delivery of the team-building program (Loughead & Bloom, 2013). In indirect interventions, the sport psychologist works with the coaching staff to create a team-building program and develop specific strategies, which are subsequently delivered and implemented with the athletes. In other words, the sport psychologist acts as a consultant for the coaching staff, who has the direct responsibility to implement the team-building protocols with their athletes (Loughead & Bloom, 2013). On the other hand, direct interventions involve the sport psychologist working directly with all members of the team (i.e., coaching staff and athletes). To this end, the sport psychologist, coaching staff, and athletes share the responsibility of creating and implementing the team-building programs. Thus, the sport psychologist is in direct contact with the athletes during program development and delivery (Loughead & Bloom, 2013).
Carron and Spink (Carron & Spink, 1993; Spink & Carron, 1993) developed and implemented an indirect team-building intervention in exercise settings. Importantly, their intervention was based on a conceptual framework that represented a linear progression of group development that included inputs, throughputs, and outputs (Carron, Spink, & Prapavessis, 1997). Specifically, group environment and group structure were the two main categories of input, which influenced the throughput category of group processes. Subsequently, group processes influenced the output, which mainly pertained to the cohesiveness of the group. Each category within the framework included a specific factor that was emphasized and targeted during the team-building intervention. For example, the group environment was targeted by enhancing the group’s distinctiveness, which reflected the extent to which the group appeared unique in comparison to other groups. Group structure mainly related to the norms and positions established within the group, while group processes included interaction, communication, and sacrifices among teammates as the most salient factors. Lastly, the output category of group cohesion included the four sub-dimensions of cohesion (i.e., ATG-T, ATG-S, GI-T, GI-S).
Using this framework, Carron and Spink conducted a set of team-building intervention studies with female exercise class participants over a 13-week period (Carron & Spink, 1993; Spink & Carron, 1993). In each study, the exercise classes under the experimental condition were led by leaders who were trained to implement team-building protocols in addition to standard exercise programs, whereas leaders in the control condition provided the standard exercise programs only. Specifically, the team-building training was delivered in four stages: introductory, conceptual, practical, and intervention (for full descriptions of the stages, see Carron et al., 1997). In the introductory stage, the authors educated the group leaders on the benefits of group cohesion such as greater self-esteem, trust, and adherence to the program, as well as more group stability. In the conceptual stage, the framework of team building was outlined to the group leaders. In this way, the group leaders were able to decide what specific factors within the framework should be targeted in their team-building program. Based on this assessment, in the practical stage, the group leaders brainstormed strategies that would enhance the specific factor. Finally, in the intervention stage, the strategies developed in the previous stage were implemented by the group leader. The four dimensions of group cohesion, as well as satisfaction with the exercise classes, were included as outcome measures. In their results, the participants in the experimental condition showed higher perceptions of ATG-T (Carron & Spink, 1993; Spink & Carron, 1993) and satisfaction with the classes (Carron & Spink, 1993), as well as adherence to the classes represented by the number of dropouts and late arrivals to each class (Spink & Carron, 1993). These results provided initial evidence for the usefulness of indirect team-building interventions.
More recently, several sport and exercise psychology researchers extended the early work by Carron and Spink (Bruner & Spink, 2010; Bruner & Spink, 2011; Newin, Bloom, & Loughead, 2008). Bruner and colleagues (Bruner & Spink, 2010; Bruner & Spink, 2011) used Carron and Spink’s model to conduct team-building interventions in school-based exercise programs. Ten exercise classes with a total of 100 adolescent (13–17 years) participants were randomized into an experimental group or a control group. The exercise classes were run three times per week over a period of eight weeks (i.e., a total of 24 sessions), each lasting approximately an hour. Following Carron and Spink’s protocols, the leaders in the control group ran a standard exercise program only, while the leaders in the experimental condition were trained to conduct team-building activities in addition to the exercise program. Their results revealed that the participants in the experimental condition reported higher task cohesion (Bruner & Spink, 2010), group task satisfaction, and session attendance (Bruner & Spink, 2011), and that the five specific factors targeted in the intervention significantly improved the prediction of task cohesion (Bruner & Spink 2010). Thus, the findings by Bruner and colleagues extended the usefulness of Carron and Spink’s four-stage model of team building to youth populations.
In sport, Newin and Colleagues (2008) conducted a team-building program with eight youth ice hockey teams. Following Carron and Spink’s (1993) model, they educated the head coaches on the benefits of team building (i.e., introductory stage), introduced the conceptual framework (i.e., conceptual stage), and developed specific activities that were designed to be engaging and challenging their athletes’ problem-solving and teamwork skills (i.e., practical stage). Then, the coaches led five activities throughout their season, which lasted approximately 30 minutes per activity (i.e., intervention stage). The authors gathered qualitative data using pre- and post-intervention reflection forms completed by coaches, observations of the activities by members of the research team, and individual semi-structured exit interviews with the coaches following the completion of the season. Among their results, coaches reported that their athletes improved their problem-solving skills, abilities to focus and to persist through challenges, and their teamwork skills. Taken together, the recent work by Bruner and colleagues (Bruner & Spink, 2010; Bruner & Spink, 2011) and Newin et al. (2008) provide evidence that Carron and Spink’s indirect team-building interventions can be beneficial under both sport and exercise contexts.
Based on his work with coaches and athletes at Penn State University, Yukelson (1997) advocated the use of a direct service approach, where the sport psychologist is in contact with the athletes during team-building interventions. Similar to the indirect approach by Carron and Spink (1993), Yukelson’s approach consisted mainly of four stages: assessment, education, brainstorm, and implementation (Loughead & Bloom, 2013; Yukelson, 1997). In the assessment stage, the sport psychologist spends time to learn about the dynamics of the organization, including its goals, needs, norms for productivity, and team atmosphere. Then, the sport psychologist educates the team on the objectives of team building and the nature of group development. Although the brainstorm stage is equivalent to Carron and Spink’s practical stage where specific team-building strategies are developed, Yukelson’s brainstorm stage involves athletes as active participants during strategy development. Finally, the strategies are implemented in the final stage. In addition to the four stages, Yukelson also described the core components that must be included in order to build a successful team. Specifically, the team-building program must promote a shared vision that encompasses the group’s overarching goals and expectations, collaborative and synergetic teamwork as a result of role clarity and acceptance among members, and individual and mutual accountability that reflect their willingness to accept responsibility for their actions and group outcomes. Further, the team must establish a positive team culture and cohesive group atmosphere where the players put the group’s interest ahead of their personal interests, a team identity that includes the team’s distinct characteristics and the extent to which the members feel proud of their membership, and open and honest communication that allows members to freely and effectively express and exchange their feelings and thoughts. Finally, the team members must be willing to provide peer helping and social support (Yukelson, 1997).
Following Yukelson’s direct approach, Voight and Callaghan (2001) conducted a team-building intervention with two NCAA women’s soccer teams. The authors conducted needs assessment for both teams that involved discussions among the coaching staff and the athletes, which led to establishing two primary objectives: team unity and performance. Based on these objectives, the consultant and the team brainstormed specific strategies to be utilized in the team-building interventions, which included individual and team goal setting, pre-performance routines, and establishing re-focusing plans, among others. These interventions were delivered in a four-day workshop during pre-season for the first team, whereas weekly team-building sessions were held for the second. In their results, self-reported intervention feedback revealed that the athletes rated the team-building program generally effective for their team unity, as well as individual and team performance.
More recently, a particular form of team-building activity that involves enhancing mutual understanding among team members has gained research attention (Dunn & Holt, 2003, 2004; Holt & Dunn, 2006; Pain & Harwood, 2009). According to Crace and Hardy (1997), team functioning can be improved when individual members go beyond understanding their own values and are able to recognize other members’ values, needs, and strengths. Similarly, Yukelson (1997) advocated the promotion of mutual understanding among teammates by open and honest communication practices. Building on this approach, Holt and Dunn delivered a pair of personal disclosure mutual sharing (PDMS) interventions, one with a male intercollegiate ice hockey team (Dunn & Holt, 2004) and another with a female high performance soccer team (Holt & Dunn, 2006). Both teams had qualified to participate in the national championship tournament at the time of the interventions. Specifically, prior to their departure to the national championship tournament, all athletes were asked to prepare a story that was personally significant in their sporting or non-sporting life. Then, the sport psychologist conducted a formal team meeting with the athletes the day before their first game of the tournament, where each athlete shared their stories with their teammates (for detailed descriptions of the intervention, see Holt & Dunn, 2006). Following the end of the season, the athletes were invited to participate in semi-structured interviews. Two separate inductive analyses of the interview data revealed that the PDMS intervention had numerous benefits that ranged from understanding self and others, to an enhanced sense of closeness and willingness to play for each other, and to feeling extremely confident in their abilities as a team (Dunn & Holt, 2004; Holt & Dunn, 2006). These results support the use of PDMS interventions, particularly with an elite group of performers who may benefit from maximizing their group functioning prior to entering a critical performance event.
Despite the encouraging results of the PDMS interventions, Holt and Dunn (2006) commented that the intervention may not be as useful at other stages of the season where the athletes’ emotional intensity and commitment are not as high, such as mid-season. As such, Pain and Harwood (2009) took a slightly different approach in their mutual sharing intervention, which involved four weekly team meetings mid-season rather than a single meeting prior to a championship tournament. Further, each meeting involved open team discussions among coaches and athletes regarding various factors related to their team functioning instead of sharing personal stories. The authors collected weekly survey data from the start to the end of the season that captured the athletes’ perceptions of their team environment and performance. Their results suggested that the athletes reported increased social cohesion, trust and confidence in teammates, as well as perceptions of team performance as a result of the intervention. Taken together, although preliminary, research evidence supports the effectiveness of team-building interventions that involve enhancing mutual understanding among the team members. More research is warranted in this regard to establish a stronger base of empirical support and to understand the various contextual factors (e.g., gender, competition level, timing of the season, length of the intervention) that may influence its effectiveness.
Team Goal Setting Approach
Although a sport psychologist may have a long list of team-building strategies to choose from, one particular strategy that seems to have the strongest empirical support is team goal setting. In fact, Martin et al.’s (2009) meta-analysis of 17 sport team-building interventions revealed that team goal setting was not only one of the most popular strategies employed, it was also one of the most effective strategies.
Based on the early work by Widmeyer and Ducharme (1997), Eys, Patterson, Loughead, and Carron (2006) introduced a three-stage team goal setting program. The program starts in stage one by explaining the rationale of the team goal setting to the athletes. Then, the athletes collectively set their team goals, following a sequence of activities that involve breaking down broad, long-term goals into more specific, short-term goals that are more readily achievable by the athletes. Specifically, the athletes first set long-term (e.g., high team standing at the end of the season) and short-term (e.g., winning three out of the next four games) outcome goals. Based on these goals, each individual athlete is then asked to determine specific performance targets (e.g., number of rebounds per game) that must be achieved in order to meet their team goals. These targets are then discussed among a subgroup of three to five players, which are then further discussed and agreed upon the team as a whole. In stage two, these performance targets are monitored on a game-by-game basis, which may involve coach feedback and/or posting the relevant statistics in a locker room. In the final stage, the sport psychologist provides ongoing feedback to the team, and the team can collectively adjust and modify their goals as needed.
An example of a team goal setting program based on the framework by Eys et al. (2006) was conducted by Senécal, Loughead, and Bloom (2008) with female high school basketball teams. In their study, eight teams with a total of 86 players were randomly assigned to either an experimental or a control condition. The experimental group was assigned the team goal-setting program described by Eys et al. over a 5-month season, whereas the teams in the control group completed measures of cohesion twice during the season without the team goal setting program. Their results showed that the teams in the experimental condition reported significantly higher perceptions of cohesion on all four dimensions than the control group at the end of the season, a difference that was not observed at the start of the season. A more in-depth analysis of their data showed that the experimental group did not change in their perceptions of cohesion over the course of their season, while the control group significantly decreased their perceptions of cohesion over the course of the season, which was attributed to a ceiling effect due to high levels of cohesion in the beginning of the season (Senécal et al., 2008). Thus, it may be concluded that a team goal setting intervention could be useful in maintaining the team’s levels of cohesion over the course of a season, which may naturally decrease otherwise. Similar to other types of team-building interventions, more research studies under various contextual elements (e.g., gender, sport, competition levels) are needed to establish a more solid basis of empirical support and external validity.
Limitations and Future Directions for Team-Building Research
While the team-building literature in sport and exercise has established useful protocols and showed some promising results in enhancing the quality of team functioning, it is also worthwhile to consider several limitations in the current literature as well as directions for future research. First, the most fundamental need within the team-building literature is that more empirical evidence is needed to support the use of team-building protocols with a variety of performance groups. For instance, Martin et al.’s (2009) meta-analysis of team-building interventions in sport was only able to identify 17 independent studies for review. Although Bruner et al.’s (2013) recent citation network and path analyses of the team-building literature identified 118 relevant articles, their review included books and book chapters, as well as populations outside sport.
Second, there is clear evidence that most team-building programs in sport have largely focused on group cohesion as an outcome variable (Martin et al., 2009). While cohesiveness of a group is an important variable for assessing and improving team functioning, and research based on cohesion has provided fruitful information, this overemphasis on cohesion “suggests that research conducted within the area of team building in sport is relatively narrow” (Bruner et al., 2013, p. 37), possibly overlooking other important individual (e.g., performance, confidence, anxiety) and team (e.g., role ambiguity, role clarity, collective efficacy) factors that may be affected by team-building interventions. McEwan and Beauchamp (2014) described in their conceptual framework of teamwork that team-building interventions may benefit from a more process-oriented approach where observable teamwork-related behaviors (e.g., goal setting, member interactions, performance monitoring) are targeted, which could “improve team functioning and effectiveness, with increased cohesion emerging over time as a by-product [emphasis added]” (p. 244). Third, in relation to the second point, future research studies may benefit from employing a more tailored approach. That is, rather than assuming that team functioning will be improved upon increased perceptions of cohesion (or any other variable), a sport psychologist may conduct team-by-team a-priori assessments to understand the specific needs of each team and employ relevant strategies. For instance, a team that needs to improve their communication practices may benefit from conducting formal team meetings to facilitate team discussions, whereas a team with low perceived levels of social cohesion may organize social events to promote positive relationships among team members.
Arrow, H., Poole, M. S., Henry, K. B., Wheelen, S., & Moreland, R. (2004). Time, change, and development: The temporal perspective on groups. Small Group Research, 35, 73–105.Find this resource:
Benson, A. J., Šiška, P., Eys, M., Priklerovád, S., & Slepičkab, P. (2016). A prospective multilevel examination of the relationship between cohesion and team performance in elite youth sport. Psychology of Sport and Exercise, 27, 39–46.Find this resource:
Bosselut, G., McLaren, C. D., Eys, M. A., & Heuzé, J. P. (2012). Reciprocity of the relationship between role ambiguity and group cohesion in youth interdependent sport. Psychology of Sport and Exercise, 13, 341–348.Find this resource:
Brawley, L. R., Carron, A. V., & Widmeyer, W. N. (1987). Assessing the cohesion of teams: Validity of the Group Environment Questionnaire. Journal of Sport Psychology, 9, 275–294.Find this resource:
Brawley, L. R., & Paskevitch, D. M. (1997). Conducting team building research in the context of sport and exercise. Journal of Applied Sport Psychology, 9, 11–40.Find this resource: