Table 4

Summary of included articles’ study design, context, method of identifying brokers and key findings about brokers
Authors, date Study design* Brokers identified by Context, settings Findings about brokers
Ahuja, G. (2000) [34] 1. Interorganisational Nonredundant contacts per total contacts Firm collaborations within the international chemicals industry Brokering structural holes between companies increases innovative output up to a point before it decreases.
2. Longitudinal, retrospective
3. Documentary data
4. Regression analyses
Aral, S. & Van Alstyne, M. (2011) [35] 1. Interpersonal Network constraint Employees from a US executive recruiting firm Brokers’ success at accessing novelty depends on their knowledge environment.
2. Cross-sectional
3. Analysis of email content
4. SNA, word mining
Balkundi, P., Barsness, Z. et al. (2009) [36] 1. Interpersonal Betweenness centrality 19 teams from across two US paper and wood-based building product plants Leaders who were brokers (high betweenness centrality) in the advice-seeking network had teams with higher team conflict and lower viability.
2. Cross-sectional
3. Paper-based survey using roster
4. SNA
Bercovitz, J. & Feldman, M. (2011) [37] 1. Interpersonal Measure of "expertise distance" between academic departments; number of ties to external networks Academic research teams from two US universities Costs are involved in coordinating diverse teams but such teams are more successful inventors.
2. Cross-sectional
3. Documentary data: invention disclosures, personnel records, patents
4. PROBIT modelling
Burt, R. (2004) [12] 1. Interpersonal Network constraint US electronics company managers Brokers accrue social capital by being able to see and express more “good ideas.”
2. Longitudinal, retrospective
3. Online survey; archival data
4. SNA; regression analyses
Colazo, J. (2010) [38] 1. Interteam Boundary-spanning activity (number of team members who work on another project per number of members in focal team) Open source software development teams Boundary spanning activity in teams was positively associated with quality but negatively associated with productivity.
2. Longitudinal, retrospective
3. Archival data on teams and project quality
4. SNA, regression analyses
Creswick, N. & Westbrook, J. (2010) [39] 1. Interpersonal Betweenness centrality Communication between ward staff of an Australian teaching hospital SNA can identify strategic people that act as brokers.
2. Case study
3. Paper-based survey using roster
4. SNA
Cummings, J. & Cross, R. (2003) [25] 1. Interpersonal Effective size 182 work groups (average 8 members) in a US Fortune 500 telecommunication firm Leaders who act as brokers ("go-betweens") within teams can cause a bottleneck in information flow that can decrease productivity.
2. Cross sectional
3. Email survey using roster
4. Regression analyses
Di Marco, M., Taylor, J. et al. (2010) [28] 1. Interpersonal Betweenness centrality Indian and US post-graduate students in two engineering project teams Nominated cultural boundary spanner (CBS) can decrease cultural based knowledge system conflicts and trigger emergent CBS.
2. Ethnographic
3. Observation over 3 days
4. SNA
Fleming, L., Mingo, S. et al. (2005) [40] 1. Interpersonal External ties (ln) 35,400 inventors across 16 East German regional innovation networks Brokers can generate innovative ideas but their presence can hamper its diffusion and use.
2. Longitudinal, retrospective
3. Archival patent data
4. Regression analyses
Hanson, D., J. Hanson, et al. (2008) [41] 1. Interpersonal Betweenness centrality 152 members of an Australian network of community groups for safety promotion Asymmetric distribution of influence: six members with high centrality and betweenness centrality.
2. Longitudinal case study, prospective
3. Paper-based survey; 3 initial waves of snowballing to identify members
4. SNA
Hargadon, A. & Sutton, R. (1997) [42] 1. Interpersonal Observation Design engineers at IDEO, a US product design firm Technology brokering involves four stages: access, acquisition, storage and retrieval.
2. Ethnographic
3. Observation, interviews
4. Grounded theory
Hawe, P. and L. Ghali (2008) [43] 1. Interpersonal Betweenness centrality Staff and teachers at a Canadian high school SNA useful tool to identify people of strategic influence (including brokers) in health promotion activities.
2. Case study
3. Paper-based survey using roster
4. SNA
Heng, H. K., W. D. McGeorge, et al. (2005) [44] 1. Interpersonal Betweenness centrality; effective size and efficiency (SH) Department managers of an Australian hospital Facility manager had high brokerage potential.
2. Case study
3. Survey using name generator
4. SNA
Lingo, E. & O'Mahony, S. (2010) [29] 1. Interpersonal Observation; assessment of tertius orientation (tertius gaudens or tertius iungens) Independent music producers in the Nashville (US) country music industry Brokerage is a process (cf. position) and both tertius orientations can be used to produce collective outcomes.
2. Ethnographic
3. Observation, interviews
4. Grounded theory
Luo, J.-D. (2005) [26] 1. Interpersonal Betweenness centrality 296 workers in two multinational technology companies in mainland China and in Taiwan Brokers ("go-betweens") in advice-seeking networks have informal power and are higher in particularist trust than others.
2. Cross-sectional
3. Survey
4. Regression analyses
Marrone, J., Tesluk, P. & Carson, J (2007) [45] 1. Interpersonal Self- and alter-assessment 190 MBA students in 31 teams in a US university consulting project Team level boundary spanning mitigates the negative cost of individual boundary spanning.
2. Cross-sectional
3. Survey
4. Hierarchical linear modelling (individuals nested within teams)
Obstfeld, D. (2005) [30] 1. Interpersonal Constraint; tertius iungens orientation Designers, engineers and managers in a US engineering division of automotive manufacturer Tertius iungens orientation, social knowledge and network density are independent predictors of involvement in innovation.
2. Ethnography, case study
3. Email survey using name generator, interviews, observation
4. Qualitative, regression analyses
Padula, G. (2008) [46] 1. Interorganisational "Shortcuts:" number of cumulative alliances to other clusters US mobile phone firms Network cohesion and brokerage ("shortcuts") synergise to produce best environment to generate and produce innovation.
2. Longitudinal, retrospective
3. Archival patent data
4. Regression analyses
Rangachari, P. (2008) [47] 1. Interpersonal Between subgroups in structural equivalence analysis Administrators and professional staff from four hospitals in New York State Brokerage across professional subgroups results in better coding performance.
3. On-line survey using roster; interviews
4. SNA; structural equivalence analyses
2. Cross-sectional
Rodan, S. & Galunic, C. (2004) [48] 1. Interpersonal Network sparseness = 1-Density Managers from a Scandinavian telecommunications company Access to heterogeneous knowledge may be more important than sparse network structures for innovative managerial performance.
2. Cross-sectional
3. Paper-based surveys using roster and one wave of snowballing to include named external contacts
4. Regression analyses
Soda, G., A. Usai, et al. (2004) [49]/ Zaheer, A. and G. Soda (2009) [50] 1. Interpersonal then aggregated to team level Network constraint TV production specialist teams from Italy Current brokerage associated with higher team performance. Past brokerage ties are not as effective as current ones.
2. Longitudinal, retrospective
3. Archival data on 501 TV
productions
4. SNA, regression analyses
Susskind, A., P. Odom-Reed, et al. (2011) [51] 1. Interpersonal Network constraint, effective size, efficiency and hierarchy Members of 11 hospitality management programs across six hotels and 11 US universities Level of brokerage was not significantly related to individual team member performance but negatively related to overall team performance.
4. SNA, regression analyses2. Cross-sectional
3. Survey using roster
Tiwana, A. (2008) [52] 1. Interpersonal "Bridging ties" extent of heterogeneity of expertise, background and skills of fellow team members 173 team members within a US internet business applications company Both strong ties and brokerage (“bridging”) ties are needed to realise knowledge integration.
2. Cross-sectional
3. Survey
4. Regression analyses

*Study design legend: 1. Level of analysis (nodes as individuals, teams or organisations); 2. Design; 3. Method of data collection 4. Method of analysis.

Long et al.

Long et al. BMC Health Services Research 2013 13:158   doi:10.1186/1472-6963-13-158

Open Data