Member communication as network structure: Relationship with task cohesion in sport

Abstract

Recent research has found that perceived information exchange between sport team members was positively associated with greater task cohesiveness. The purpose of the current research (two studies) was to extend these findings and test if (a) the proportion of teammates with whom an athlete exchanges information and (b) the proportion of members in the group collectively who exchange information were associated with perceived task cohesion. Participants (N = 205) from intact teams in Study 1 identified the proportion of team members with whom they regularly exchanged information using a nomination procedure, and a one-way ANOVA was used to differentiate between three groups (i.e. athletes who identified that they interacted with a larger, moderate, and smaller proportion of teammates) in terms of task cohesion. Results revealed a significant group effect, partial η² = .16, p < .001. As predicted, those interacting with a larger proportion of teammates reported greater task cohesion than those interacting with a moderate and smaller proportion of teammates (all ps < .05). Using a hypothetical vignette design, a sample of team sport athletes (N = 177) in Study 2 read a team description and were presented with a corresponding communication network graphic that varied in overall network density. As hypothesised, those exposed to the team portrayed with higher network density reported higher anticipated task cohesion than those exposed to the lower network density team description vignette (p < .001).

Keywords:

• social networks
• communication
• information exchange
• cohesion

Notes

1 These data were screened for potential sex differences in the information exchange network size/cohesion relationship. Given that the pattern of ANOVA results were the same across the male and female portion of the sample, the data of males and females were collapsed. Team size also was considered as a potential control variable given the potential for larger teams to deflate proportion scores. Given that team size only slightly differed within this sample (70% of respondents with a team size of 10–12 athletes) and the correlation between team size and proportion of teammates nominated was not significant (r = .04, p = .55), we did not control for team size in order to preserve power.

2 A comparison of means suggested no differences in demographic characteristics (i.e. age, sex, competitive level, years of sport experience, starting status, team size, and whether participants were currently competing on their team) between those who correctly recalled the degree to which members exchanged information and those who recalled incorrectly. As such, there appeared to be no systematic reasons that determined sensitivity to the manipulation.

3 To ensure that hypothetical team size did not account for differences in perceptions of task cohesion, a series of additional independent samples t-tests were performed. For the higher density condition, participants viewing smaller (n = 25) versus larger sized teams (n = 35) did not significantly differ in perceptions of GI-T (p = .68). For the lower density condition, those viewing smaller (n = 27) versus larger sized teams (n = 37) also did not significantly differ in perceptions of GI-T (p = .13). All manipulation-check measures, including cooperative communication, followed the same pattern of non-significant differences. As such, scores were combined across the team-size conditions vignettes.

Additional information

Funding

This research was supported by a Doctoral Scholarship to the lead author from the Social Sciences and Humanities Research Council of Canada [grant number 752-2014-2655].