Mapping 50 Years of Small Group Research Through Small Group Research

Cluster 1: Team Performance

A great deal of work on small groups and teams incorporates performance as a primary dependent variable and most of it (unless specialized into its own cluster) finds its home in Cluster 1 (light blue). This cluster contains a set of terms that have occupied group researchers’ attention for a long time: teamwork, team performance, leadership, personality, team development, team processes, and innovation. It is a good sign that these lines of research have used their longevity to identify each other and integrate across boundaries. It is also not surprising that this cluster sits at the center of all four networks across periods. This cluster is integrated with other clusters, although, as we will describe below, some terms and clusters are more strongly connected to its content than others.

Cluster 2: Virtual Teams

The two top terms in Cluster 2 (dark pink) are “virtual teams” and “computer-mediated-communication (CMC).” These lay the groundwork for the rest of the cluster examining diversity, creativity, and transactive memory, which have been applied to the study of virtual teams in work and communication environments (e.g., Eisenberg et al., 2019; Gilson et al., 2015; Han & Beyerlein, 2016; Martins & Shalley, 2011; O’Leary & Mortensen, 2010).

Cluster 3: Cohesion and Communication

Two of the terms that have differentiated their own literature are team cohesion and communication, which are the two terms most linked to Cluster 3 (dark green). These terms have been linked to group processes and development over time as well as conceptualizations of groups as social networks (e.g., Aarts et al., 2016; Carron & Brawley, 2012; Rogelberg & Rumery, 1996; Rovio et al., 2009; Van Swol & Kane, 2019).

Cluster 4: Learning Dynamics

Like Clusters 2 and 3, the top two terms in Cluster 4 (lime green) set the tone for the cluster: “team learning” and “time.” This work incorporates a good deal of material on shared leadership (see Koeslag-Kreunen et al., 2018 for a review) as well as work on team interaction and interdependence as antecedents to learning over time (e.g., Ishak & Williams, 2017; Määttä et al., 2012; Zoethout et al., 2017).

Cluster 5: Group Decision-Making

The group decision-making literature, driven by work in both psychology and communication, is the focus of Cluster 5 (orange). It incorporates terms such as “decision-making,” “group decision-making,” “information sharing,” and “hidden profile.” Lu et al. (2012) and Mesmer-Magnus and DeChurch (2009) provide reviews of the hidden profile and group decision-making literature.

Cluster 6: Collective Efficacy and Emotional Intelligence

Cluster 6 (teal) contains work on collective efficacy and emotional intelligence. For example, Little and Madigan’s (1997) article on collective efficacy in manufacturing teams, and Chang et al.’s (2012) article on the impact of emotional intelligence on small groups are represented in this cluster. It also contains work that combines these two topics like that by Hjertø and Paulsen (2016).

Cluster 7: Creativity

Cluster 7 (light pink) contains articles on creativity. For example, McLeod et al. (1996) provide evidence that small groups containing ethnically diverse members come up with better ideas than ethnically homogeneous small groups. In addition, this cluster contains several articles published as part of SGR’s special issue on meetings in April, 2012, which are often a context for creativity.

Cluster 8: Interdisciplinary Collaboration

Cluster 8 (rose) identified separately from Cluster 4, and contains only four terms: group performance, collaboration, interdisciplinary, and group cohesion. The articles in it indicate a focus on interdisciplinary collaboration and cohesion (e.g., Chang & Bordia, 2001; Evans & Dion, 1991; Stokes, 1983).

Cluster 9: Group Counseling

Cluster 9 (yellow) contains the terms group dynamics, groups processes, and group counseling, which come together in papers on group counseling (e.g., Hornsey et al., 2007; Kivlighan & Mullison, 1988; Page et al., 1980).

Cluster 10: Consensus

Cluster 10 (turquoise) contains only two terms, pertaining to consensus and the measurement of group environments (e.g., DeStephen & Hirokawa, 1988; Schyns, 2006). These are linked because consensus continues to be an important issue in environmental perceptions (Whitton & Fletcher, 2014). Consensus also continues to be an important topic in the group decision-making and top management team literatures (Bragaw & Misangyi, in press).

Phase 1 (1970–1981)

Table 5 gives raw counts of terms used during this period. Figure 1 shows the network structure of communication between clusters from 1970 to 1981. Analyzing Table 5 and Figure 1 together sheds light on several important factors in the development of the study of small groups. The first insight from Table 5 is that this early period of group research saw a focus on group longitudinal development. Terms like “group process,” “group dynamics,” “time,” and “group development” dominated the titles and abstracts of articles published during this period. Because of this, group processes that naturally occur over time and impact group development such as “communication,” and “conflict” also saw focus. These terms also reflect the editorial focus of William Fawcett Hill. Terms such as “training” and “group counseling” are indicative of work on T-groups (or training groups) as therapeutic interventions found in the social work literature (e.g., Fisher & Werbel, 1979; McIntire, 1973). In fact, during this period, the term “training” appeared in title and abstracts 3 times more than any other term.

OPEN IN VIEWER

Table 5.

Top 20 Keyword Occurrences in the SGR Corpus 1970–2019.

1970–1981		1982–1991		1992–2010		2011–2019
Keyword	Count	Keyword	Count	Keyword	Count	Keyword	Count
Training	32	Training	39	Performance	251	Performance	135
Communication	16	Group Process	38	Communication	110	Communication	45
Leadership	11	Communication	31	Time	82	Time	40
Time	9	Leadership	30	Cohesion	78	Cohesion	32
Group Counseling	8	Performance	27	Group Process	74	Leadership	25
Group Process	8	Group Development	21	Decision-Making	60	Interdisciplinary	25
Performance	6	Conflict	20	Leadership	59	Teamwork	24
Group Dynamics	5	Decision-Making	19	Work Group	59	Diversity	24
Cohesion	5	Time	18	Conflict	57	Conflict	19
Work Groups	4	Cohesion	16	Satisfaction	44	Group Process	19
Conflict	3	Consensus	12	Group Development	29	Collaboration	18
Personality	3	Satisfaction	11	Training	27	Decision-Making	17
Group Development	3	Personality	8	Group Dynamics	26	Meetings	14
Consensus	2	Meetings	8	Computer-Mediated Communication	25	Work Groups	13
Interaction Analysis	2	Cooperation	5	Personality	24	Training	13
Interdisciplinary	1	Work Groups	4	Teamwork	24	Virtual Teams	13
Decision-Making	1	Interdisciplinary	4	Consensus	21	Satisfaction	12
Meetings	1	Group Dynamics	3	Diversity	21	Team Learning	12
Satisfaction	1	Creativity	3	Meetings	16	Creativity	11
Innovation	1	Innovation	2	Brainstorming	15	Computer-Mediated Communication	10
Trust	1	Trust	2	Virtual Teams	13	Faultlines	8

Note. Keywords from 1970 to 2002 were identified in titles and abstracts since keywords did not exist until 2003. We also excluded the words “group” and “team” from these lists.

Still, overall, the terms we observe in Phase 1 are broadly descriptive of positivist group functioning focusing on specific cause–effect relationships (McGrath et al., 2000). This broad terminology relates to the cursory study of groups as interdisciplinary phenomena during this period. Although phenomenologically groups were always understood to contain elements relevant to multiple disciplines, group research grew out of work in social psychology (McGrath et al., 2000) until that waned beginning in the late 1960s leading other fields to pick up the important study of group functioning (Levine & Moreland, 1990). This is evidenced in Figure 1 by the lack of communication between several prevalent research communities. While work in the Team Performance cluster is linked to early conflict work (Cluster 2) and work on cohesion (Cluster 3) and meetings (Cluster 7), and work within this cluster is linked in examining how leadership and team composition affect team processes such as communication which ultimately impact team performance, we also observe a great deal of early work on group dynamics and decision-making isolated from this central cluster.

Importantly, these broad initial terms lay the groundwork for the deeper exploration of group functioning that occurs in later phases where terms such as “virtual team,” “social network,” “transactive memory,” and “collective efficacy” come to prominence. During this first period, the groundwork for the Virtual Teams cluster can be seen in work on conflict and diversity (e.g., Blanchard, 1975; Pood, 1980). In addition, the theoretical bases of collective states such as collective efficacy and transactive memory were developed during or shortly following this period (e.g., Bandura, 1977; Wegner et al., 1985).

Phase 2 (1982–1991)

Figure 2 depicts the network structure and communication between clusters from 1982 to 1991. Analysis of it, combined with Table 5, show a developing and more interdisciplinary field of small group research. Group processes and development are still a major focus, which is beneficial given the naturally dynamic and longitudinal nature of groups (Waller et al., 2016). This is indicated by the continued heavy use of terms such as “communication,” “conflict,” “group process,” “group development,” and “time.”

Second, during this period we clearly see more integrated communication between clusters. This aligns with the increased consideration of groups as interdisciplinary phenomenon during this period (Weiss & Hoegl, 2015) and the increased understanding that similar inputs and mechanisms (such as composition and emergent states) span group settings (Mathieu et al., 2017). For example, work on conflict was drawn into work on group decision-making (Pendell, 1990; Wiiteman, 1991). Work on group dynamics based in the group counseling literature (Cluster 9) was also being considered in central work on leadership and performance (Cluster 1, for example, Counselman, 1991; Forsyth et al., 1985; Zamarripa & Krueger, 1983). In addition, work on decision-making and conflict became more integrated with psychological work on cohesion and personality, and began to integrate with management work on leadership. We also see an increased focus on satisfaction (Cluster 5) as an outcome of group interaction and integration of this work with broader work on leadership (Cluster 1) and time (Cluster 4).

In addition, we see several new terms come to prominence during this period indicative of early work underlying current established topics in small group research. Work on brainstorming and creativity (e.g., Burton, 1987; Comadena, 1984), team learning (e.g., Krayer, 1988; Kuriloff et al., 1988), and cooperation (e.g., Stahelski & Tsukuda, 1990; Tyerman & Spencer, 1983) all began from 1982 to 1991. Finally, it is worth noting that as William Fawcett Hill left the editorship, so did the journal’s focus on psychoanalytic groups. Group counseling went from the fifth most-used term from 1970 to 1981 to the 21st most-used from 1982 to 1991. The addition of Fred Massarik as editor likewise saw a rise in the discussion of performance (4 to 21 uses) commiserate with his business school background, and focus on organizational groups. This focus continued through the next two time periods.

Phase 3 (1992–2010)

In 1990, Charles Garvin and Richard Kettner-Polley began the first editorship of the modern version of SGR, and the longest of any time period. Because of the length of this period and coinciding developments in the small group literature (Mathieu et al., 2017; Weiss & Hoegl, 2015) we focus on four major takeaways from it, supported by Figure 3 and Table 5.

First, the rise of performance as an outcome of group interaction must be noted. Although performance was always considered a key outcome of group functioning, in previous periods this consideration was at parity with other processes such as communication, training, and leadership. From 1982 to 1991 performance was the fifth most-used term, accounting for just 3% of the terms used. From 1992 to 2010 performance was a clear first. It was used more than twice as much as the second most-used term (communication), accounting for 9% of overall term usage. This seems to indicate that during this period the group literature shifted to focus on how processes (and emergent states and compositional factors) affect the ability to attain group goals.

Second, the impact of technological phenomena on the study of small groups rose significantly during this period. The study of virtual teams (Cluster 2, including both “virtual teams” and “CMC”) became much more sophisticated and integrated with other areas involving trust, conflict, diversity, brainstorming, and group creativity. This makes sense since the phenomenon of virtual teams itself became more complex. This observation also has implications for the general development of novel terms and subfields pertaining to the study of groups, which we address in the discussion section.

Third, from 1992 to 2010 many new terms emerged at the periphery of the keyword network. These terms are widely recognized as important in the small groups literature; examples include “transactive memory,” “hidden profile,” “emotional intelligence,” “collective efficacy,” “faultlines,” and “social networks.” Although the theoretical and empirical roots of many of these terms such as social networks, collective efficacy, transactive memory, and hidden profile, were developed prior to 1992, their use was not prevalent enough to warrant consideration in earlier keyword networks. Considering the pattern of this addition in reference to Figures 1 and 2 indicates that new terms needed to be established as distinct constructs (while still being connected to their host literatures) before scholars could integrate them with extant parts of group literature (Kwon & Adler, 2014).

Fourth, in Figure 3 we see areas of integration between previously disparate clusters. For example, virtual teams is much more connected to the Team Performance (Cluster 1), Creativity (Cluster 7), and Cohesion and Communication (Cluster 3) clusters, indicating that the virtual team’s literature draws from a wide and well-developed base to understand this important phenomena. Creativity is also communicating much more with Group Decision-Making (Cluster 5), integrating these two traditional outcomes of group interaction. This is important because it signals a recognition in the creativity literature that groups provide a logical environment for the generation and ultimate selection of creative ideas because the natural diversity of backgrounds and knowledge inherent in groups provides a larger knowledge base from which to pull ideas (Amabile, 1996; Emich & Vincent, 2020; Woodman et al., 1993). Furthermore, the naturally disparate perspectives found in groups, inherent in each individual member, allows for the novel recombination of ideas through communication patterns often studied in the group decision-making literature, which underlies how creative ideas are generated (West et al., 2003). Thus, the integration of these literatures provides the path to a more nuanced understanding of how group communication patterns influence the collective generation and selection of ideas. Finally, Learning Dynamics (Cluster 4) is integrated to a much greater degree into the central Team Performance cluster (Cluster 1). This signals an important recognition that as group environments became more complex and dynamic, the ability of teams to adapt and learn became integral to their ability to continually perform at a high level (Bell et al., 2012). This integration is also notable because the intertwined nature of not only learning and performance, but also motivation, forced deeper examination of the theoretical underpinnings of these constructs including regulatory theories (Kanfer et al., 2008; Karoly, 1993), information processing models (Hinsz et al., 1997; Huber, 1991; Walsh, 1995), and macrocognition (Cacciabue & Hollnagel, 1995; Fiore et al., 2010). This examination has clear implications for the integrated study of learning, motivation, and performance as learning, development, and adaptation become more important to navigating complex and dynamic team environments. This development also has implications for exploring how the combination of regulatory, information processing, and macrocognitive processes lead to the emergence of important team attitudes and processing ultimately resulting in team performance and knowledge outcomes (Bell et al., 2012).

It is also important to note that the increasingly broad focus on different aspects of SGR during this period is likely related to Charles Garvin’s choice to add Richard Kettner-Polley as a co-editor. Their complementary backgrounds in social work and social psychology/organizational behavior allowed for the incorporation of work from a wide range of disciplines relevant to small groups.

Phase 4 (2011–2019)

When Joann Keyton and Aaron Brower took over editorship of SGR in 2009, they continued the tradition of Charles Garvin and Richard Kettner-Polley, balancing work from different communities of small group research and strengthening interdisciplinary ties between clusters. Brower’s background in social work and Keyton’s background in communication were similarly complementary to Garvin and Kettner-Polleys’. In addition, Keyton’s background in communication facilitated the integration of communication work with dominant work in management and psychology during this period (Norder et al., 2018), particularly broadly applicable work on communication mediums and technology. Evidence of this is displayed in Figure 4 and Table 3.

To this point, the first takeaway from this period is that focus on the Virtual Teams cluster (Cluster 2) intensified in line with the increased use of CMC and virtual teams throughout organizations and broader society (Tannenbaum et al., 2012). For example, in line with the broad applicability of the phenomenon, work on virtual teams further expanded into that on decision-making (McLeod, 2013), group diversity (Martins & Shalley, 2011), cultural diversity (Han & Beyerlein, 2016), training (Gilson et al., 2013), faultlines (Chiu & Staples, 2013), and group dynamics (Handke et al., 2019). This integration signals the promise and peril of virtual teams. They are advantageous because they can accomplish tasks that collocated teams cannot. Virtual teams can assemble deeper pools of knowledge by combining disparate experts. They can incorporate 24-hr productivity, lower costs, and share knowledge broadly. However, these advantages come at a cost. Virtual teams rely on technologies more, which poses a direct cost for organizations and requires team members to adopt to the new technology. Relying on CMC alone, instead of complementing it with face-to-face communication, brings social costs as well. For example, it is difficult for leaders to share appropriate information and develop important team attitudes such as psychological safety and shared mental models in virtual teams (Hoch & Kozlowski, 2014; Zaccaro & Bader, 2003). It is also more difficult to resolve conflict, motivate group members, and build trust, especially across cultural differences (Dulebohn & Hoch, 2017). Our observed massive integration of research topics and communities with virtual teams signals the wealth of knowledge that has been applied to addressing this difficulty. Importantly, this integration signals sustained research opportunity as newer technologies become more prevalent, and in fact ubiquitous, for group communication. Despite the progress made in understanding virtual teams, there is work to be done exploring how factors such as technology, globalization, and trust building influence virtual team effectiveness (Gilson et al., 2015).

Second, this aligned with an increased focus on faultlines and diversity as explanations of group processes. This focus included discussions of measurement techniques (Meyer et al., 2014) and effect moderators (Meyer et al., 2011). This work also integrated with more established topics, such as leadership (Schölmerich et al., 2016) and virtual teams (McLeod, 2013). Importantly, this integration marks a continued effort on the part of group researchers to deal with groups as complex, multilevel, dynamic systems (Mathieu et al., 2019) composed of subjective members (Emich & Lu, 2017). This is no small endeavor. The diversity and faultlines literature has begun to address it by adapting assumptions of configurational emergence to account for the multiple unique attributes of individual group members. For example, traditionally, if one was to consider group member age they may use a mean or standard deviation. In addition, if they were to consider sex or ethnicity, they may use a ratio. However, the faultlines literature allows the concurrent consideration of these attributes in predicting important team processes, most notably subgrouping and conflict (Meyer et al., 2014). Still, the faultlines literature is only an initial signal of the wealth of knowledge that can be gained by considering the concurrent impact of group member attributes on group processes and outcomes, just as demographic variables represent a small portion of individual differences that may affect group functioning. Continued integration of this work with a broader set of communities will allow researchers to examine the importance of the concurrent variation of any number of group member attributes, whether those attributes be demographic, cognitive, behavioral, or affective/motivational. For example, this basic premise could be used to explore questions, such as: Does it matter when high status members begin task conflict as compared with low status members? Does it matter when proactive team members are also conscientious? Or, does it matter when members with a better understanding of a transactive memory system are more empathetic?

Third, qualitative research was more heavily embraced and used to investigate phenomena such as groupthink (Lee, 2019), team learning (Zoethout et al., 2017), meetings (Köhler et al., 2012), and collaborative learning (Isohätälä et al., 2020), among others. This marks an important development as the incorporation of qualitative analyses into this traditionally quantitative field can allow deeper insight into the nuances surrounding the complexity of group interaction. For example, by analyzing recordings of teacher teams, Zoethout et al. (2017) were able to track how group learning processes emerged from group conversations, particularly the extent to which group learning processes change when members act on each other’s reasoning.

Finally, few new terms were incorporated into the small group lexicon during this period, despite heavy patterns of increased interaction between topics and clusters; however, shared leadership stands as an exception. During this period, the study of shared leadership developed in the Learning Dynamics cluster (Cluster 4) and integrated with work on the longitudinal study of teams and team effectiveness and learning (e.g., Koeslag-Kreunen et al., 2018; Wang et al., 2017). Research on shared leadership has also been integrated with research in the virtual teams cluster on diversity (Xu et al., 2019). Importantly, this work fits into the tradition of combining work on small groups from multiple disciplines to take a collective process approach to leadership, which has more traditionally been considered to flow top-down from a single source (Pearce & Conger, 2002)