A Social Network Analysis of the EMAC Annual Conferences 2000-2010
1. Introduction
In recent years there has been a growing interest in the nature of scientific collaboration.
Collaboration between researchers play an important role in scientific development in
general and consequently, numerous studies have showed an increase in the number of
co-authored research papers (Laband and Tollison, 2000; Moody, 2004). Members of
social institutions of science over a broad range of disciplines co-author research papers
together creating social networks of researchers. These social networks play a crucial
role in scientific growth as they work to share and generate new knowledge through
social interaction (Crane, 1972). The manner in which authors influence each other
when collaborating on joint publications is important in the process of producing
knowledge (Bourdieu, 2005).
Social Network Analysis provides a mean for interpreting and measuring relationships
between a number of social entities, such as people, groups and organisations. The
emphasis on relationships is an important supplement to standard social and behavioural
research, which is mainly concerned with attributes of the social entities (Wasserman
and Faust, 1994). In Social Network Analysis the attributes of the individual actors are
not essential, the focus is on the structure of their relationships and how the structure of
linkages affect individual actors and their relationships. The structure of a network
provides insights into network activities and how knowledge is generated and shared
within the network.
An efficient way for scientific researchers to exchange and bring forth knowledge is
through collaboration in specialist organisations such as the European Marketing
Academy. The structure in networks like this is often hidden because of its informal
network characteristics.
To date little attention has been paid to the application of Social Network Analysis in a
conference setting. The objective of the study at hand is to employ Social Network
Analysis to analyse the research collaboration within a specific academic group over
time, namely the European Marketing Academy. It should be seen as a contribution to
the existing literature in the field of Social Network Analysis as it provides a review of
previous literature relating Social Network Analysis and co-authorship. Furthermore, it
1
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
adds to the limited number of empirical studies applying Social Network Analysis in a
conference setting.
The current study will focus on the structure of the EMAC community by analysing the
way in which researchers of this specific network collaborate. The patterns and
regularities in the choice of collaboration partners of members of the EMAC
community will be unveiled to discover if general assumptions can be made concerning
the manner in which these researchers work together.
1.1 Problem Statement
A considerable amount of research can be ascribed to members of the European
Marketing Academy over the past 35 years. The vast majority of scientific papers
presented at the academy´s annual conference have been co-authored by two or more
researchers. The way in which these academics choose to work together is characterised
by the absence of any formal hierarchy.
The creation of knowledge is often a joint process where patterns and regularities in the
way in which scholars work together emerge. These patterns and regularities provide
valuable insight into how knowledge is created and shared within groups of researchers
and can also give ideas to what powers such networks. By using Social Network
Analysis as a sociological approach for analysing patterns of relationships and
interactions between researchers, the underlying social structure of a scientific
collaboration network can be discovered.
In order to understand the research community of the European Marketing Academy
better it is relevant to gain insights into the morphology of the network in terms of
clustering to establish the grounds on which the researchers in the EMAC community
select their collaboration partners.
The aim of the study at hand is to demonstrate if patterns and regularities in the way in
which authors work together exist within the European Marketing Academy by
performing a Social Network Analysis. It will seek to answer the following research
questions:
2
A Social Network Analysis of the EMAC Annual Conferences 2000-2010

What are the characteristics of the structure of the European Marketing
Academy scientific collaboration network?

Which factors influence the choice of collaboration partners in the European
Marketing Academy scientific collaboration network?
1.2 Structure
In order to give a background for the current study, the section “Background” will
provide a description of the concept of Social Network Analysis and give an
introduction to the European Marketing Academy. The literature review will then
review previous literature linking Social Network Analysis and co-authorship,
presenting the main findings of these studies. This is followed by the methodology
section, which explains the methods applied, the unit of analysis and the constructs
utilised in Social Network Analysis. The section then continues with a description of
how the data set was constructed and finishes with a detailed desription of the measures
employed in Social Network Analysis. The results section will present the findings of
the study. This section is followed by a discussion of the results.The study at hand will
then review the limitations and give suggestions for further research in the field. The
study will conclude with an assessment of the main findings.
2. Background
2.1 Social Network Analysis
Social Network Analysis concerns the comprehension of the connections among social
actors and the consequences of these connections. It reveals a structure of linkages,
within which actors are embedded. Actors are described by their relations, not by their
attributes and the relations are just as fundamental as the actors that they connect. As a
tool Social Network Analysis has its roots in the social sciences. The central concepts of
relation, network and structure originates from a number of disciplines within the social
sciences such as sociology and anthropology. Social Network Analysis has gained wide
use in disciplines as diverse as economics, marketing and industrial engineering. This
can, in part, be ascribed to the fact that Social Network Analysis provides insight into
aspects of the political, economic and social structural environment (Wasserman and
Faust, 1994).
3
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
In general, Social Network Analysis can be defined as a structured way of analysing
relationships within groups (Cross, Borgatti and Parker, 2002) by “providing a rich and
systematic means of assessing informal networks by mapping and analysing
relationships among people, teams, departments or even entire organisations” (Cross,
Parker, Prusak and Borgatti, 2001: 103).
In order to asseess the characteristics of a network Social Network Analysis utilises a
unique set of diagnostic tools. The set of methods and analytic concepts used in Social
Network Analysis have been developed over the past 50 years as an inherent part of
progress made in social theory, empirical research and formal mathematics and statistics
(Wasserman and Faust, 1994).
Social Network Analysis has a broad range of application possibilities. One such is the
analysis of informal networks of academics collaborating on research papers. As this
type of network is characterised by the absence of any formal hierarchy, a Social
Network Analysis could reveal patterns and regularities in the way in which academics
work together to generate knowledge. Furthermore, it could disclose the structure that
shape the creation of knowledge within a given field of research (Vidgen, Henneberg
and Naudé, 2007).
2.2 The European Marketing Academy
The European Marketing Academy (EMAC) was founded in 1975 as a professional
society for individuals involved or interested in marketing theory and research. The
objective of the society is to act as the core of a communication network for distributing
information and promoting international exchange in the field of marketing. The
Academy currently has more than 1000 members from more than 57 different countries
and is the largest European community of marketing scholars (http://www.emaconline.org/r/printPage.asp?iID=IHGMD).
The majority of EMAC members joined the Academy after hearing about it from a
colleague. The social aspect of being a member of EMAC is very important to
especially the younger members, who consider network opportunities and career
development a significant reason for joining the community. Particularly the posibility
4
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
of participating in the annual conferences attracts many researchers to the Academy
(Wong, 2010).
The highlight of the year for the European Marketing Academy is the annual
conference held at a major university or scientific institute in Europe. The conference
serves as a forum where members exchange ideas and present and discuss research
projects
in
all
areas
of
marketing
(http://www.emac-
online.org/r/printPage.asp?iID=IIKFF). It is the scientific collaboration of authors
presenting research papers at the annual EMAC conference that forms the basis of the
study at hand.
One of the core activities of EMAC is the publication of the A level journal
International Journal of Research in Marketing (IJRM). The journal has served as
EMAC´s flagship journal for more than 25 years. The number of submissions has seen
an increase in recent years in addition to the enhancement of the quality of articles
submitted to the journal (Dekimpe, 2010). Besides IJRM, the Academy publishes The
Chronicle, which is a bi-annual publication informing members of ongoing activities in
the marketing discipline. Additionally, the Academy publishes an online newsletter in
order to keep members up to date with the activities of the Academy (http://www.emaconline.org/r/printPage.asp?iID=IHGMD).
3. Literature Review
In recent years, there has been an increasing amount of literature on Social Network
Analysis. This large and growing body of literature has, among other things, applied
Social Network Analysis as a tool for conducting citation analyses (see for instance
Zinkhan, Roth and Saxton, 1992 and Carter, Leuschner and Rogers, 2007).
Additionally, a large volume of published studies employ Social Network Analysis to
reveal patterns and regularities in the way in which scholars work together by focusing
on co-authorship in published research papers (e.g. Newman, 2001a; Newman, 2001b;
Barabási, Jeong, Néda, Ravasz, Schubert and Vicsek, 2002; Liu, Bollen, Nelson and
Van de Sompel, 2005). So far, however, only limited attention has been paid to the
application of Social Network Analysis in a conference setting.
5
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Co-authorship is the most formal expression of intellectual collaboration in scientific
research (Acedo, Barroso, Casanueva and Galán, 2006). Social Network Analysis
pertains to the comprehension of the linkages among social entities and the implications
of these linkages. As a tool Social Network Analysis has its origin in the social sciences.
As mentioned above the central concepts of relation, network and structure originates
from a number of disciplines within the social sciences such as sociology and
anthropology (Wasserman and Faust, 1994). Social Network Analysis has gained wide
use in a broad area of disciplines (Eaton, Ward, Kumar and Reingen, 1999; Laband and
Tollison, 2000, Newman, 2001b; Barabási et al., 2002; Morlacchi, Wilkinson and
Young, 2005; Vidgen et al., 2007; Henneberg, Jiang, Naudé and Ormrod, 2009). The
tendency for co-authorship has been increasing across almost all scientific disciplines.
Various explanations for this trend has been put forward. First, the cost of equipment
for conducting scientific experiments in the natural sciences is rather high and this
might encourage collaboration. In the social sciences, researchers are not as dependent
on labs, however, large-scale data collection induces collaboration as researchers can
share the heavy workload. Second, scientific collaboration provides better opportunities
for specialisation and division of labour as it is far more efficient to bring in a new
scholar than to learn new material oneself. This often results in specialists being brought
in to conduct the analyses (Laban and Tollison, 2000; Moody, 2004).
In a case study of the International Marketing and Purchasing (IMP) Group 1984-1999,
Morlacchi et al. (2005) used Social Network Analysis to analyse research collaboration,
in the form of co-authorship, within this specific academic group over time. The focus
of the study was on the people relationships inside the IMP Group based on coauthorship reported in the proceedings of the annual IMP conferences. In their study,
Morlacchi et al. (2005) selected a sample from the population of interest which
comprised authors who had attended 3 or more annual IMP conferences and who
therefore had been involved in 3 or more papers. This was done to center the study on
more active members of the group who would attend the annual conference on a regular
basis and present papers that would add to the generation of new knowledge in the field.
A serious weakness to this methodology, however, is that imposing such a restriction
will cause the study to give an inaccurate picture of the structure and operation of the
IMP Group, as scientists who have produced less than 3 research papers obviously also
contributes to the creation of new knowledge within the field. By omitting certain
6
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
scientists, the study will not convey the true nature of the research collaboration
network of this specific academic group.
In contrast to the methodology of Morlacchi et al. (2005), Henneberg et al. (2009), in
investigating the same academic group (i.e. the IMP Group, 1984-2006), utilised all
papers presented at the annual IMP conference to give a coherent picture of the IMP
community. In doing so an artificial boundary to the network was created. Vidgen et al.
(2007) points out that excluding all co-authored papers not presented at the conference
is a methodological disadvantage, a disadvantage they, nevertheless, are willing to
accept.
Newman (2001b) conducted a comparative study of co-authorship focusing on three
fields of research, namely physics, biomedical research and computer science. In order
to make valid comparisons of collaboration patterns in the three different fields a fairly
short time window of five years was used. This implies that the collaboration network
was kept static during the study and that time evolution of the network was not
examined. This research approach differs from that of Barabási et al. (2005). In their
different, but complementary, approach to collaboration networks Barabási et al. (2005)
mapped the electronic database of journals in mathematics and neuro-science over an 8year period in order to derive the dynamic and structural mechanisms that regulate the
evolution of the collaboration networks in the two fields. Similarly, Morlacchi et al.
(2005), Vidgen et al. (2007) and Henneberg et al. (2009) conducted longitudinal studies
of time periods of minimum 13 years to infer the evolution of the complex systems
investigated.
A number of studies have found that the distribution of the number of co-authored
research papers does not follow a normal or bell shaped distribution. Instead it follows a
power law distribution where few scholars contribute a large number of co-authored
papers and where many authors contribute a small number of co-authored papers
indicated by the long tail of the distribution (Lotka, 1926; Newman, 2001b; Barabási et
al., 2002; Morlacchi et al., 2005).
Previous research have found the research communities under examination to not be
fully connected due to the existence of clusters. In other words, the networks contained
7
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
subgroups which were not connected to other subgroups through paths between authors
(Vidgen et al., 2007; Henneberg et al., 2009). However, despite the lack of links
between a number of subgroups the main component of the IMP Group network showed
a relatively robust network structure facilitating information dissemination through the
network (Henneberg et al., 2009). In comparison, the network of ICT researchers
presented by Vidgen et al. (2007) exhibited a quite delicate nature where the removal of
a few key authors would lead to a division of the main component into subcomponents.
Morlacchi et al. (2005) demonstrated a tendency for geographical grouping within the
IMP Group. This trend becomes evident as disconnected subgroups emerge in the
network containing researchers from only one country. A possible explanation to the
appearance of such subgroups could be the language barrier. These findings are
supported by Liu et al. (2005) who, in their study on co-authorship in the digital library
research community, found collaboration between authors from different countries to be
a mere 7 %. Recent evidence suggests a degree of preferential attachment within social
networks of collaborating auhtors as researchers tend to be organised around institutions
and collaborate closely within specific clusters and groups of interest (Eaton et al.,
1999; Barabási et al., 2002; Liu et al., 2005).
Although research communities grow over time given that more academics join in to
express their intellectual ideas, they often form “small worlds”, by which the average
separation between the individual researchers is shorter than would appear from random
data (Vidgen et al., 2007). In their study of the IMP Group, Morlacchi et al. (2005)
found the network to be highly clustered, where some researchers collaborate
extensively within the network and play a key role in linking different parts of the
network. This result is supported by Newman (2001b), who reported that networks of
collaborating scholars form “small worlds” as the typical distances between scholars
within the networks are small. Barabási et al. (2002) showed that even though networks
grow over time, the average distance between any two scholars actually decreases over
time, lending support to the idea of networks forming “small worlds”. Similarly, Moody
(2004) reports that the pattern of co-authorship within the discipline of sociology shows
a “steadily growing cohesive core” (2004: 235). However, in contrast to the results
mentioned above Vidgen et al. (2007), in their study on co-authorship in the European
Conference on Information Systems, 1993-2005, found that the main component of this
8
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
specific research community displayed few properties of a “small world”. Components
with 15 or more actors did indeed demonstrate “small world” characteristics. Still, if
one of the scholars in one of these smaller components co-authors a scientific paper
with a researcher from the main component, the smaller component will be absorbed
into the main component. This surprising result is interestingly enough backed by the
results of the study conducted by Henneberg et al. (2009). They found there to be a core
within the IMP Group where there were fewer small clusters working on their own. This
would indicate that the IMP Group displays more of a “large world” disposition, in
which it is more probable that the actors are connected through co-publishing, and thus
expressing the same opinions. This result is surprising because the earlier study of the
same academic group conducted by Morlacchi et al. (2005) provided results supporting
the “small world” view.
An additional interesting topic presented in earlier literature concerns the presence of
so-called “invisible colleges” within social networks. According to Crane (1972), social
interaction between researchers play a role in scientific knowledge development. She
sees the growth of scientific knowledge as a sort of diffusion process where central
actors in a social system, who have adopted an idea, influence other actors who have yet
to adopt this new idea. These networks of highly productive scientists form “invisible
colleges” which hold a key role in disseminating knowledge throughout fields of
research. These informal communication networks or “invisible colleges” work to link
separate groups of scholars within a research area and to provide coherence and
direction within the field (Crane, 1972; Morlacchi et al. 2005).
The presence of “invisible colleges” within social networks were found in a study on
co-author relationships and author productivity in consumer behaviour research
conducted by Eaton et al. (1999). The structure of the research community under
investigation revealed trace characteristics of an “invisible college”. It showed the
relation of collaboration among highly productive authors and their importance in
securing the intellectual structure of the field. Henneberg et al. (2009) found strong
cohesion between groups of centrally important actors in the IMP Group indicating
continuous collaboration over time and thus lending support to the notion of “invisible
colleges” as centres of knowledge creation. The findings of Morlacchi et al. (2005)
corroborates the results presented by Henneberg et al. (2009) as they also found an
9
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
“invisible college” to be present within the IMP Group connecting subgroups of
researchers from different countries.
A limitation to a number of previously conducted studies on co-authorship is that they
only consider one aspect of research collaboration (Eaton et al., 1999; Newman, 2001b;
Barabási et al., 2002; Liu et al., 2005; Morlacchi et al., 2005; Henneberg et al., 2009).
Vidgen et al. (2007) extended their research to include two types of relationships,
namely co-authorship and panel membership. Nevertheless, the authors still name the
limited types of relationships investigated as the chief weakness to their study. Focusing
on co-authorship in a conference setting, Morlacchi et al. (2005), Vidgen et al. (2007)
and Henneberg et al. (2009) neglect to include co-auhtorship in books and journal
articles outside of the conference setting, which is a disadvantage as these are also
important ways for co-authoring researchers to publish their ideas.
In his study on the strength of weak ties between actors in a network, Granovetter
(1973) argued that the stronger the tie connecting two individual actors, the more
similar these actors will be. According to his argument this also implies that if strong
ties connect actor A to actor B and actor A to actor C, then actor B and actor C will both
be similar to actor A and thus similar to each other. This will in turn increase the
likelihood that actor B and actor C will forge a friendship once introduced. Furthermore,
in stressing the importance of weak ties, he reasoned that information can reach a larger
number of people and cover greater social distance when passed through weak ties
rather than strong ties (Granovetter, 1973). This is due to the fact that weak ties connect
individuals in otherwise separate clusters whereas strong ties tend to share the same
connections (Burt, 1995). That is, the people an individual is weakly tied to move in
different circles and will therefore be able to spread information to a wider network
(Granovetter, 1973).
4. Methodology
The present study will follow the methodology used by Vidgen et al. (2007) and
Henneberg et al. (2009) and will analyse the pattern of co-authorship in research papers
presented at the European Marketing Academy annual conferences between 2000 and
10
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
2010. It will, however, differ from the two abovementioned research papers, in that it
will analyse the years of the conference individually and not cumulatively.
The unit of analysis is the co-authored, scientific conference paper, i.e. all research
papers presented at the EMAC conferences 2000-2010 with two or more authors. This
means that the type of relationship studied is co-authorship. By focusing the study on
co-authorship rather than citation, emphasis is placed on the social relationships of a
network as it must be assumed that authors co-authoring a scientific research paper have
an actual social relationship. In accordance with Vidgen et al. (2007) and Henneberg et
al. (2009), the analysis of the 11 years of EMAC conferences will adopt a longitudinal
view of discreet data points (i.e. the annual conferences).
The network analysis of the study at hand aims at mapping the development of the
network of the EMAC community over the past 11 years. The focus of the study will
therefore be on the years 2000 to 2010 neglecting conferences held prior to this. In order
to give a coherent picture of the development of the network in the time period under
investigation the current study will concentrate on the years 2001, 2004, 2007 and 2010.
Social Network Analysis makes use of two constructs: Nodes and linkages. Nodes
characterise data points in the network, which in the current study are authors. Linkages
represent connections between nodes, which in the present analysis attest to the fact that
two or more authors have co-authored a research paper and presented it at one of the
EMAC conferences in the period 2000 to 2010 (Henneberg et al., 2009). The study at
hand will use the terms node, actor, scholar, researcher and author interchangeably. The
terms linkage, relationship, connection and co-authorship are also used interchangeably.
In analysing social networks the number of linkages between actors characterises the
interconnectedness of the network in addition to the idea of tie strength. This implies
that if actor A has co-authored research with actor B and with actor C, then a linkage
between actor A and actor B and a linkage between actor A and actor C exists. The
strength of the ties between the actors can be assessed when looking at the number of
research papers the actors have co-authored. If actor A has written three papers with
actor B and only two papers with actor C, the strength (value) of the tie between actor A
and actor B is stronger than that of actor A and actor C (Henneberg et al., 2009).
11
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
For the Social Network Analysis in question, the UCINET programme was used to
conduct the analyses. The NetDraw programme was utilised to visualise the analyses
performed in UCINET (Borgatti et al., 1999).
4.1 The Data Set
The data source for the Social Network Analysis in the present study was retrieved
manually from the European Marketing Academy conference proceedings 2000-2010.
The study at hand limits its coverage to competing research papers presented at the
conferences neglecting poster- and special sessions. Papers marked “withdrawn” in the
2010 conference proceeding were also excluded from the data set.
Authors were identified by surname, which could lead to problems when names are
inconsistent over time, when errors in the conference proceedings make distinguishing
between first name and surname difficult or when two or more authors share the same
surname. In order to avoid these problems a search was performed after entering each
author name into the data set. If the search showed that a specific author name was
already present in the data set, an inspection of the conference proceedings was done
manually to investigate whether the specific author was already present in the data set.
Authors sharing the same surname were given a code name so as to avoid confusion
between individual authors (see Appendix 1 for list of code names). Through the
aforementioned search and manual inspection of the conference proceedings
inconsistencies in author name were also identified and corrected. Discrepancies in
method of listing author names in the conference proceedings challenges the validity of
the present study. In a number of the conference proceedings authors were listed by
surname followed by first name, other proceedings listed authors by first name followed
by surname and some applied a combination of the beforementioned methods. The
method of searching the data set and manually inspecting the conference proceedings
eliminated several of the discrepancies.
In order to categorise the measures described below it is important to quantify the
properties of the relationships under investigation. The relationships (linkages) in the
EMAC data set are non-directional and valued. In a directional relationship the tie
between two actors originates from a point and has a destination, i.e. the tie is directed
12
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
from one actor to the other. This property would be relevant if the study set out to
measure for instance friendship. However, co-authorship between two authors does not
have a direction so the relationship is non-directional. Dichotomous relationships are
labeled as either present or absent for each set of actors. This type of relationship makes
no attempt in measuring the strength of the relationship. The EMAC data set is valued
as it is possible to measure the strength of the ties between actors in the network. This is
done by observing the frequency of authors co-authoring research papers together. A
high number of joint research papers will indicate a stronger tie between the actors
(Wasserman and Faust, 1994; Henneberg et al., 2009).
4.2 Measures
A number of social network measures are available for evaluating the characteristics of
a collaboration network. As mentioned above, the linkages in the EMAC data set are
non-directional and valued. These properties will influence the way in which a number
of the measures in Social Network Analysis are defined. The following section will
provide an introduction to the key term “components” and the measures will then each
be considered in turn.
4.2.1 Components
An important characteristic of a graph is whether it is connected or not. In order for a
graph to be connected a path must exist between all pairs of nodes in the graph. This is,
however, not always the case. Often a graph is partitioned into two or more subsets
where no connection between the subsets exist. These subset are refered to as
components. This means that all actors in a component can reach each other but they
can not reach members of other components in the graph. The main component of a
network is the component with the largest number of connected actors (Wassermann
and Faust, 1994). Many analyses require that all the nodes are connected and it will
therefore in many cases be relevant to extract the main component in order to perform
the analysis on a fully connected network.
13
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
4.2.2 Connection and Distance
4.2.2.1 Density
Density can be defined as “the ratio of the number of lines present to the maximum
possible that could arise” (Wasserman and Faust, 1994: 143). In other words, the
density of a network can tell how many of all possible ties are present in a network.
This will in turn give insights into the speed at which information is diffused among
members of the network and how cohesive the structure of the network is. A wellconnected network will be better at capitalising its resources and bringing forth a
diverse range of perspectives for problem solving. The density of a network also
provides insights into the role of the individual actors in a network. Some networks
might be composed of quite similar actors while other networks can contain an elite of
central actors with many ties and a larger group of actors with a small number of ties
(Hanneman and Riddle, 2005). The variation in the degree of connectivity of actors in a
network became evident in a number of studies as these saw the emergence of so-called
“small worlds” and “invisible colleges” (Crane, 1972; Eaton et al., 1999; Newman,
2001b; Barabási et al., 2002; Moody, 2004; Morlacchi et al., 2005; Henneberg et al.,
2009).
4.2.2.2 Geodesic Distance
The geodesic distance of a network can be defined as “the length of the shortest path
between two nodes” (Knoke and Yang, 2008: 60). The term is used to describe more
complex characteristics of an individual´s position in the network and the structure of
the network as a whole (Hanneman and Riddle, 2005). A problem arises when a
network is not fully connected, here it is not possible to define the exact geodesic
distance between all actors (Hanneman and Riddle, 2005). To overcome this problem
the main component of the network can be extracted and the analysis can be performed
on this specific part of the network. This will not yield a completely accurate result as
the analysis is not performed on the cumulative data set, however, important properties
of the individual actors and the structure of the main component will become evident.
Previous research has shown that despite of a network growing over time, the geodesic
distance between two actors actually decreases over time which supports the notion of
networks forming “small worlds” (Barabási et al., 2002; Moody, 2004; Morlacchi et al.,
2005).
14
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
4.2.2.3 Diameter
The diameter of a connected graph can be defined as “the length of the largest geodesic
distance between any pair of nodes” (Wasserman and Faust, 1994: 111). The term
diameter gives information on the size of the network. It shows how many steps are
necessary to get from one side of the network to the other side of it (Hanneman and
Riddle, 2005). As with the geodesic distance, the diameter of a network can not be
calculated accurately if the network is not fully connected. It is therefore necessary to
extract the main component and then perform the analysis. If a network´s diameter is 5
or lower, it is said to exhibit “small world” properties (Henneberg et al., 2009). Earlier
studies on co-authorship in a conference setting have shown that the main component of
these research communities are not “small worlds”, on the contrary they display “large
world” properties with diameters reaching 25 and 31 (Vidgen et al., 2007; Henneberg et
al., 2009).
4.2.3 Centrality Measures
The centrality measures in Social Network Analysis are primarily used to identify the
most important actors in a social network. These measures seek to describe the location
of central actors in networks as the most prominent actors often take up strategic
positions within networks (Wasserman and Faust, 1994).
4.2.3.1 Actor Degree Centrality
Actor degree centrality measures “the extent to which a node connects to all other nodes
in a social network” (Knoke and Yang, 2008: 63). An actor who has a high degree
centrality level takes up a hub position in the network as he/she will be in direct contact
with or adjacent to many other actors. These central actors will be highly visible within
the network and other actors will acknowledge their central location in the network as
leading channels of relational information (Wasserman and Faust, 1994). With a larger
number of connections, central actors have better opportunities for communcation
because they have a greater selection of actors to communicate with. This means that
central actors are less dependent on other individual actors and this gives them more
social power within the network (Hanneman and Riddle, 2005).
15
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
4.2.3.2 Actor Closeness Centrality
Actor closeness centrality indicates “how near a node is to the other nodes in a social
network” (Knoke and Yang, 2008: 65), both directly and indirectly (Wasserman and
Faust, 1994). This measure focuses on how close an individual actor is to the other
actors in the network. It reflects the ability of an actor to access information through the
network and the actor is central if he/she can communicate with all other members of
the network rather quickly (Wasserman and Faust, 1994). An actor´s closeness
centrality is a function of the shortest path between each individual (also known as the
geodesic distance, see above) and every other person in the network (Knoke and Yang,
2008). The closeness centrality of an actor is inversely connected to distance which
means that when nodes grow farther apart in distance from each other, its closeness
centrality will decrease as the shortest path between members of the network becomes
larger (Wasserman and Faust, 1994). Actors who are able to reach other actors at shorter
path lengths enjoy a favourable position in the network. They serve as a reference point
and their opinions are heard by a large number of actors, hence they have more social
power within the network (Hanneman and Riddle, 2005).
4.2.3.3 Actor Betweenness Centrality
Actor betweenness centrality relates “the extent to which other actors lie on the shortest
path between pairs of actors in the network” (Knoke and Yang, 2008: 66). An actor will
then be central if he/she lies on the geodesic distance between a large number of other
actors. This measure takes into account the connectivity of the node´s neighbours and
reflects the number of individuals an actor connects indirectly through their direct links
(Wasserman and Faust, 1994). The central actors are in a position where they can
influence the interactions between two nonadjacent actors as they can control the flow
of information between other members of the network, this position naturally gives the
actor more social power within the network (Hanneman and Riddle, 2005).
The three actor centrality measures have been utilised extensively in the literature.
Focusing on Social Network Analysis applied to a conference setting, several studies
identified key researchers who dominated the centrality measures. These central actors
play important hub roles in the networks which they are embedded in (Morlacchi et al.,
2005; Vidgen et al., 2007; Henneberg et al., 2009).
16
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
4.2.3.4 Eigenvector Centrality
The actor closeness centrality measure described above indicates how close actors are to
each other in social networks. When dealing with larger and more complex networks,
however, this measure can be slightly misleading. The closeness measure of two actors
can assess the two actors as rather similar when in reality one actor is more central than
the other because he/she is able to reach more of the network with the same amount of
energy. Eigenvector centrality evaluates the importance of a node in a network in terms
of the global structure of the network and therefore pays little attention to patterns that
are more local. This is done in an effort to locate more central actors (Hanneman and
Riddle, 2005). This means that if an actor has a high eigenvector centrality score he/she
is connected to many other actors who in turn are also well-connected. Having a high
eigenvector centrality score implies that the actor is more likely to receive ideas such as
new research methods and hear of new terms and concepts within the collaboration
network (Henneberg et al., 2009).
4.2.3.5 Flow Betweenness Centrality
Flow betweenness centrality builds on the term actor betweenness centrality. Now and
then the geodesic path between two actors may be blocked by a hesitant actor who is
unwilling to pass on information between the two actors. A longer and less efficient
pathway may exist between the two actors which they will make use of in order to
bypass the reluctant actor. The flow betweenness centrality measure assumes that actors
will use all pathways that connect them. It measures how involved an actor is in the
flows between all other pairs of actors (Hanneman and Riddle, 2005). An additional
quality of the flow betweenness centrality measure is that it considers the value of
relations in a network, i.e. it separates weak ties from strong ties (Henneberg et al.,
2009). According to Henneberg et al. (2009) flow betweenness centrality can be defined
as “a measure of the extent to which the flow between other pairs of actors in the
network would be reduced if a particular actor were removed” (Henneberg et al., 2009:
41). This means that the measure can be used to see what will happen to a collaboration
network if a key researcher retires or otherwise leaves the network.
4.2.3.6 Reach Centrality
Reach centrality can be defined as “the percentage of nodes in the network that the focal
node can reach in a given number of steps” (Henneberg et al., 2009: 35). This measure
17
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
is another approach to identifying how close an actor is to other actors in a network. By
asking what part of all other actors an actor can reach in one step, two steps, three steps
etc. it can reveal how easily an actor can access information from other members of a
network (Hanneman and Riddle, 2005).
4.2.4 Structural Holes
The concept of structural holes implies that the less close knit a researcher´s co-authors
are, the more a researcher connects researchers that are otherwise not connected
(Morlacchi et al. 2005). Structural holes were first put forward by Burt (1992) who
defined the concept as “the separation between nonredundant contacts. Nonredundant
contacts are connected by a structural hole. A structural hole is a relationship of
nonredundancy between two actors” (Burt, 1995: 18). The presence of structural holes
may indicate a positional advantage or disadvantage for individuals embedded in a
network. In occupying a nonredundant connection (a structural hole) an actor can play a
key role in linking different parts of the network and thereby facilitate both the diffusion
of new ideas, research methods, theories and concepts throughout the network, and the
development of the network as a whole (Hanneman and Riddle, 2005; Morlacchi et al.,
2005). In order for an actor to be effective, he/she must distinguish between primary and
secondary contacts and then focus on maintaining relationships with the primary
contacts, who serve as “ports of access to clusters of people beyond” (Burt, 1995: 21).
By focusing on relationships with primary contacts, that is, those who provide access to
new clusters, instead of relationships with contacts who copy access to already existing
clusters (i.e. they are structurally equivalent, see more below) an actor can increase
his/her power within the network (Vidgen et al., 2007). The presence of structural holes
in networks has been demonstrated in a number of previous studies (Morlacchi et al.,
2005; Vidgen et al., 2007).
4.2.5 Cohesive Subgroups
Moving away from the way in which individual actors are connected, a macro
perspective is adopted in order to focus on the social structures within which individual
actors are embedded. Cohesive subgroups can be defined as “subsets of actors among
whom there are relatively strong, direct, intense, frequent, or positive ties” (Wasserman
and Faust, 1994: 249). In valued networks such as EMAC, a cohesive subgroup is one
18
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
where ties between actors have a high value. When evaluating cohesive subgroups a
threshold value, c, is considered for the value of ties within the subgroup. The threshold,
which ranges from 0 to c – 1, can be used to identify more or less cohesive subgroups
(Wasserman and Faust, 1994).
A disadvantage to actors in a network forming strong, cohesive subgroups is the risk of
“groupthink” occurring within the network. In a collaboration network of scientific
researchers, whose interest is to generate new knowledge and bring forth new research
methods, terms and concepts, the occurrence of “groupthink” will limit the number of
new ideas brought into the subgroup, and hence severely moderate the generation of
new knowledge within the field of research (Granovetter, 1973; Wasserman and Faust,
1994; Vidgen et al., 2009).
4.2.5.1 Cliques
A clique is a subgroup in a network where actors are more closely tied to one another
than to other actors in the network. At level c a clique can be defined as “a subgraph in
which the ties between all pairs of actors have values of c or greater, and there is no
other actor outside the clique who also has ties of strength c or greater to all actors in
the clique” (Wasserman and Faust, 1994: 278-279). Dividing a network into cliques of
actors can yield important information on how a network as a whole is likely to behave.
It can provide insights into where conflicts are likely to happen and if diffusion of new
theories will run smoothely through the network (Hanneman and Riddle, 2005).
The concept of cliques in Social Network Analysis has received criticism for being a
rather strict method for identifying cohesive subgroups. This is because the absence of
just one single tie can hinder a subgraph from being a clique (Knoke and Yang, 2008).
4.2.5.2 Structural Equivalence
In order to make inferences about patterns of relations among actors in a network it can
be useful to group together actors who are similar. This is done to predict social
behaviour of actors and social structure of networks. In categorising actors into sets of
actors who are equivalent it is possible to describe what makes these actors similar and
what makes the specific category they belong to different from other categories of
actors. The focus here is not the attributes of the individual actors but the similarities of
the patterns of relations among these actors, which translates into the social role and the
social position an actor can possess in a network (Hanneman and Riddle, 2005).
19
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Wasserman and Faust (1994) gives the following definition for structural equivalence in
a valued network like EMAC: “For two actors to be structurally equivalent on a valued
relation they must have ties with identical values to and from identical other actors”
(Wasserman and Faust, 1994: 359). In other words, actors who are structurally
equivalent hold identical positions in the structure of a social network.
As with the term cliques described above, structural equivalence is defined rather
strictly. The definition can be relaxed and it is then possible to measure how close two
actors come to being perfectly structurally equivalent. The relaxed definition implies
that “actors are approximately structurally equivalent if they have the same pattern of
ties to and from other actors” (Wasserman and Faust, 1994: 360).
4.2.5.3 Cluster Analysis
Cluster analysis divides actors into subgroups in which members are perfectly or
approximately structurally equivalent and is used to see how closely knit researchers are
within a given network. Three underlying criteria exist for establishing clusters: Single
link, average link and complete link. In order to place actors in the same position
(cluster) at a particular level of structural equivalence, a threshold value is decided upon
which serves a boundary between clusters (Knoke and Yang, 2008). Complete link
cluster analysis generates clusters where all pairs of actors are no less similar (no more
different) than the threshold value decided upon. This type of cluster analysis
outperforms alternative methods as it produces more homogeneous and stable clusters
(Wasserman and Faust, 1994). Most actors take up positions in local neighbourhoods
where the majority of actors are connected to each other, i.e. they are clustered into
local neighbourhoods. This means that the density in local neighbourhoods of large
networks have a tendency to be much higher than expected from random data of the
same size (Hanneman and Riddle, 2005).
Clustering within a network is expressed by it displaying “small world” properties.
Several studies have reported this phenomenon (Watts and Strogatz 1998; Newman
2001b; Barabási et al., 2002; Moody, 2004; Morlacchi et al., 2005).
5. Results
This section will present the results of the analyses. First, a number of basic
characteristics of the cumulative network will be displayed. Second, the main
20
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
characteristics of connection and distance will be shown. Third, the main components of
the individual years will be illustrated. Finally, the results of the more exhaustive
analyses of the four years, which are the focus of the study at hand, will be presented.
In the current study the distance measure geodesic distance will provide similar insights
to the network as the density and the diameter. Consequently, this specific measure will
not receive further attention. The measures concerning cohesive subgroups are not
relevant to the approach of the study at hand as they will cover the same aspects as the
centrality meaures and the division of the main components according to affiliation and
track. They will therefore not be investigated further.
5.1 The EMAC Network
The EMAC network is not fully connected. A number of components exist in the
network for which no connection between authors in one component and authors in
another component is present. This applies to both the cumulative network and the
individual years.
Year
No. of actors in
network
No. of actors in main
component
2000-2010
4043
2640
Table 1: Basic characteristics of the cumulative network 2000-2010.
As shown in Table 1 above more than half of all actors in the EMAC network are in the
main component indicating that researchers tend to collaborate across different
disciplines and topics within the field of marketing.
5.2 Connection and Distance
The table below presents some of the main characteristics of connection and distance
for the EMAC conferences under investigation.
21
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
No. of
Year Conference actors in
No. of
Density
Diameter
actors
of
of
in main
main
main
network component component component
2000 29th EMAC
342
17
0.0772
5
2001 30th EMAC
356
11
0.2000
4
2002 31st EMAC
532
15
0.4381
2
2003 32nd EMAC
427
13
0.1154
4
2004 33rd EMAC
622
25
0.2292
5
2005 34th EMAC
790
17
0.1213
4
2006 35th EMAC
859
25
0.0900
5
2007 36th EMAC
916
24
0.0797
4
2008 37th EMAC
763
15
0.1429
4
2009 38th EMAC
838
21
0.0738
6
2010 39th EMAC
893
37
0.0541
6
Table 2: Main characteristics of connection and distance for the EMAC conferences 2000-2010.
Table 2 reveals a general increase in the number of actors in the network, i.e. in the
number of researchers participating in the EMAC conferences 2000-2010 with coauthored papers. While the number of actors in the network has displayed a general
increase, the density of the main components has fluctuated over the 11 year period. The
density of a network is zero if no connection is present between any of the nodes in the
network. On the other hand, if every single node is connected to every other node, the
density is one (Henneberg et al., 2009). 2002 represents the year with the best connected
network in the time period as 43.81 % of all possible ties are present. 2010 is the most
sparsely connected network as only 5.41 % of all possible ties are present. The diameter
provides information about how fast information is likely to diffuse throughout a
network. The years 2000 to 2008 all have diameters of 5 or less, this means that these
years display “small world” properties. On the other hand, the years 2009 and 2010 both
have a diameter of 6 which points to less tightly knit networks.
22
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
5.3 Main Components 2000-2010
In the following the main components of 2000-2010 will be visualised. The number on
each link in the diagrams represents the strength of the tie between two actors, i.e. the
number of papers they have co-authored together.
Figure 1: Main component for the EMAC Conference 2000.
The main component for 2000 is presented in Figure 1. It reveals a somewhat scattered
group of researchers grouped around Grønhaug and Hogg with Bruce occupying a hub
role as he links two subgroups that would otherwise not be connected. The strength of
the ties between the individual researchers shows that the actors in this specific main
component has not co-authored more than one paper with each other.
23
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Figure 2: Main component for the EMAC Conference 2001.
Figure 2 illustrates the main component for 2001. It is slightly more dense than that of
2000 and identifies De Wulf and Odekerken-Schröder as the researchers who the
component is grouped around. Van Kenhove and De Wulf hold hub positions and link
two subgroups to the main component that would otherwise be disconnected. The
majority of the researchers in this specific main component have only co-authored one
research paper with each other, however, De Wulf and Odekerken-Schröder have coauthored three papers together.
Figure 3: Main component for the EMAC Conference 2002.
24
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
The main component for year 2002 is shown in Figure 3 above. It displays a solid
network with a cohesive and robust structure. Blesa represents an outlier as this specific
researcher has only co-authored with one of the actors in the network. Each member has
only written one research paper with every actor he/she is connected to.
Figure 4: Main component for the EMAC Conference 2003.
Figure 4 displays the main component for 2003. It reveals a sparsely connected network
where De Valck takes up a hub role as this specific researcher links the two parts of the
component. Again the number of research papers the members of this component has
produced with actors they are connected to equals one.
25
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Figure 5: Main component for the EMAC Conference 2004.
It can be seen in Figure 5 that the main component for 2004 is rather dense indicating a
well connected network with a robust structure. The general trend of not co-authoring
more than one research paper with other actors continues for the main component for
2004. Hooley, Greenley and Möller, however, have all co-authored two research papers
and serve as the links between two groups in the component that would otherwise not be
connected.
Figure 6 below shows the main component for 2005. It reveals a somewhat compact
network grouped around Wetzels and De Ruyter, with Wetzels occupying a hub
position as he links two otherwise disconnected parts of the component. This main
component differs from the main components of previous years as some actors have coauthored several research papers together. Again Wetzels and De Ruyter stand out as
they have co-authored four papers together. De Ruyter has, additionally, co-authored
two papers with Schepers.
26
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Figure 6: Main component for the EMAC Conference 2005.
Figure 7: Main component for the EMAC Conference 2006.
27
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Figure 8: Main component for the EMAC Conference 2007.
Figure 7 above illustrates the main component for 2006. It presents a rather cohesive
network with a compact center. Möller and Bloemer hold hub positions in the
component as they join two parts of the network to the center of the component that, if
they were not present, would be disconnected from the main component. The majority
of the tie strengths are one following the trend of earlier years. Möller and Matear have,
however, co-authored two research papers together and Matear has, additionally written
two papers with Hooley.
The main component for 2007 is displayed in Figure 8 above. It identifies two groups of
researchers linked by the collaboration of Wetzels and Stokburger-Sauer. Again the
majority of the tie strengths are one following the trend exhibited in previous years.
Three sets of actors, however, have collaborated on two research papers. These are:
Stokburger-Sauer and Bauer; Lages and Queiroga; and Wetzels and Van Birgelen.
28
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Figure 9: Main component for the EMAC Conference 2008.
Figure 9 above shows the main component for 2008. It presents a somewhat cohesive
network where Herrmanan and Tomczak act as hubs in connecting the groups of the
component. Once again the majority of the strength of the ties in the network are equal
to one, except the tie linking Tomczak and Mühlmeier who have collaborated twice in
2008.
Representing the main component for 2009 is Figure 10 below. It reveals a network
which to some degree is scattered. The collaboration between Völckner, Becker and
Ringle links a small group of researchers to the larger part of the network. Furthermore,
the collaboration between Brito and Carvalho connects another small group of actors to
the more compact part of the network. All actors in 2009 have only co-authored one
research paper with other actors to whom they are connected and all tie strengths are
therefore equal to one.
29
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Figure 10: Main component for the EMAC Conference 2009.
Figure 11: Main component for the EMAC Conference 2010.
Figure 11 above presents the main component for the final year of analysis, namely
2010. It reveals a relatively cohesive network with a comparatively robust structure.
Sattler and Hennig-Thurau serve as hubs as they link two relatively large groups to the
30
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
more compact part of the network. As in the previous years the majority of the tie
strengths are one, except the tie between Erfgen and Sattler, and the tie between Sattler
and Völckner which both equal two.
5.4 EMAC Conference 2001
In the current section the results of the more detailed analysis of the EMAC conference
of 2001 will be presented.
5.4.1 Affiliation and Track
Affiliation key:
Ghent University, Belgium
Maastricht University, The Netherlands
Figure 12: Main component for 2001 with affiliation.
Figure 12 above illustrates the main component for 2001 according to affiliation. It can
be seen that the vast majority of the researchers are affiliated with Ghent University in
Beligium pointing to strong geographical clustering within the main component for
2001. Further analysis of the main component for 2001 revealed that although the
researchers are affiliated with the same university the tracks (see Appendix 2 for list of
tracks) within which they co-author scientific research papers are diverse. This is
demonstrated in Figure 13 below. When examining Figure 13 it becomes evident that
31
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
some researchers tend to work across tracks within the field of marketing. Especially De
Wulf and Odekerken-Schröder distinguish themselves as they have co-authored within
four and three tracks, respectively. The network, however, does not exhibit track-based
clustering as each actor has only co-authored one scientific research paper within each
track.
Track key:
Relationship Marketing
Consumer Behaviour
Retailing, Channel Management and Logistics
Advertising, Promotion and Marketing Communication
New Technologies and E-Marketing
Figure 13: Main component for 2001 according to track.
5.4.2 Central Actors
The three most central actors in the main component were selected for further analysis
by assessing the neighbourhood size of all actors in the main component. Henneberg et
al. (2009) refer to neighbourhood size as the number of other actors with whom an actor
has links. The three actors with the largest neighbourhood were then chosen.
32
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
According to Table 3 below, De Wulf is the most active researcher in the main
component for 2001. Actor degree centrality identifies De Wulf as a highly visible actor
within the network. This is also reflected in the actor closeness centrality measure,
which characterises De Wulf as the researcher who is closest to other actors in the
component. The actor betweenness centrality measure takes into account the indirect
ties of an actor. It reveals De Wulf and Van Kenhove to be the leaders in the network as
they hold the highest score on this specific measure. These findings correspond to the
diagram in Figure 2 in which De Wulf and Van Kenhove are characterised as a central
actors.
Actor
Actor
Actor
Actor
Neighbourhood degree closeness betweenness
size
centrality centrality
centrality
De Wulf
7
8
76.923
Van Kenhove
6
6
71.429
Odekerken-Schröder
5
6
66.667
Table 3: Actor centrality measures for the main component of 2001.
21
21
5
Table 4 below provides additional centrality measures for the main component of 2001.
The eigenvector centrality measure locates De Wulf to be a central actor within the
network as this actor is connected to many other actors who, in turn, are also well
connected. A somewhat surprising finding is that Odekerken-Schröder has a higher
score than Van Kenhove in this measure. However, when examining Figure 2 it
becomes evident that the connections of Odekerken-Schröder are better connected than
the actors which Van Kenhove is connected to, hence the higher eigenvector centrality.
De Wulf also holds the highest score on the flow betweenness centrality measure. This
means that the flow in the main component of 2001 would be creatly reduced if this
specific researcher were to leave the network. Van Kenhove is also identified as an
actor with high flow betweenness centrality score. Again an examination of Figure 2
justifies this high value as it can be seen that Van Kenhove holds a hub position in the
network. The reach centrality measure indicates that the main component for 2001 is
rather dense as all three central researchers can reach the entire network in two steps.
Finally, the concept of structural holes shows the degree to which a researcher connects
researchers that are otherwise not connected. Again De Wulf and Van Kenhove hold
high values as they occupy hub positions in the network. This is also demontrated in
33
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Figure 2 where De Wulf links two researchers to the more compact group of researchers
and Van Kenhove connects three researchers to the group.
Flow
Reach centrality
Eigenvector betweenness
Structural
centrality
centrality
two-step
three-step
hole
Actor
De Wulf
75.217
47.667
1.00
Van Kenhove
52.360
47.000
1.00
Odekerken-Schröder
69.095
14.667
1.00
Table 4: Additional centrality measures for the main component of 2001.
1.00
1.00
1.00
5.000
4.167
2.667
5.4.3 Additional Component Analysis
The following results refer to the cumulative network for 2001. Table 5 below shows
the main component together with the second and third largest components in the
cumulative network for 2001. The density of the components are relatively high
indicating a comparatively well connected network. This is also reflected in the
diameter of the components. As none of the diameters exceed 5, they all display “small
world” properties.
Component
No. of actors
Density
Diameter
1
11
0.2000
4
2
10
0.1556
3
3
7
0.2679
2
Table 5: Main co-authorship component characteristics for 2001.
34
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Figure 14 below illustrates the second largest component of 2001 according to
affiliation. The component reveals a relatively strong geographical clustering around
Kingston Business School, England.
Affiliation key:
Kingston Business School, England
Casio Electronics Ltd.
University of South Australia, Australia
Massey University, New Zealand
London Business School, England
Figure 14: Second component for 2001 with affiliation.
35
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Additional analysis of the second largest component of 2001 shows a minor tendency
towards track-based clustering around East. East is the most active researcher within
this component, having co-authored two papers in one track and one in another track.
These findings are presented in Figure 15 below.
Track key:
Product and Brand Management
Modeling and Forecasting
Consumer Behaviour
Retailing, Channel Management and Logistics
Figure 15: Second component for 2001 according to track.
36
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
The third component for 2001 with affiliation is displayed in Figure 16 below. It reveals
very strong geographical clustering as all members of the component are affiliated with
the same institution, namely Athens University of Economics and Business, Greece
Affiliation key:
Athens University of Economics and Business, Greece
Figure 16: Third component for 2001 with affiliation.
37
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Track key:
Marketing Strategy and Leadership
Retailing, Channel Management and Logistics
Consumer Behaviour
Service Marketing
Figure 17: Third component for 2001 according to track.
Figure 17 above, shows the third component for 2001 according to track. The network
does not exhibit track-based clustering as each individual author has only co-authored
one scientific research paper within each track. Papastathopoulou is the most active
node in the component as this specific researcher has collaborated on papers in all four
tracks represented in the component.
5.5 EMAC Conference 2004
In this section the more exhaustive analysis of the EMAC conference of 2004 will be
presented.
38
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
5.5.1 Affiliation and Track
Affiliation key:
Helsinki School of Economics, Finland
Aston University, England
Budapest University of Economic Science and Public Administration, Hungary
Monash University, Australia
University of Limerick, Ireland
Univerza V Mariboru, Slovenia
City University of Hong Kong
Wielkopolska Business School – Poznañ University, Poland
ALBA – Athens Laboratory of Business Administration, Greece
University of Innsbruck, Germany
Cardiff University, Cardiff Business School, Wales
University of Wales Aberystwyth, School of Management and Business, Wales
Stern School, NYU, USA
Maastricht University, The Netherlands
University of Otago, New Zealand
39
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
University of Helsinki, Finland
Loughborough University, England
Figure 18: Main component for 2004 with affiliation.
The main component for 2004 according to affiliation is displayed in Figure 18 above.
It can be seen that a large number of universities and business schools are represented in
the diagram. This indicates very little geographical clustering within the component.
Additional analysis of the main component for 2004 revealed extremely strong trackbased clustering. This is illustrated in Figure 19 below. This result, however, is not
surprising since the majority of the researchers in the main component have
collaborated on one scientific research paper.
Figure 19: Main component for 2004 according to track.
5.5.2 Central Actors
As can be seen in Table 6 below Greenley, Hooley and Möller have been identified as
the actors in the main component of 2004 who have the largest neighbourhood. In fact,
for all measures Greenley, Hooley and Möller have the same score. This means that the
three actors hold identical positions in the network. These finding are reflected in Figure
5 above.
40
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Actor
Actor
Actor
Actor
Neighbourhood degree closeness betweenness
size
centrality centrality
centrality
Greenley
19
19
82.759
Hooley
19
19
82.759
Möller
19
19
82.759
Table 6: Actor centrality measures for the main component of 2004.
Actor
24
24
24
Flow
Reach centrality
Eigenvector betweenness
Structural
centrality
centrality
two-step
three-step
hole
Greenley
36.752
24
1.00
Hooley
36.752
24
1.00
Möller
36.752
24
1.00
Table 7: Additional centrality measures for the main component of 2004.
1.00
1.00
1.00
6.474
6.474
6.474
5.5.3 Additional Component Analysis
The succeeding results refer to the cumulative network for 2004. Table 8 below displays
the main component together with the second and third largest components in the
cumulative network for 2004. While the density of the main component is
comparatively high indicating a somewhat well connected network, the two smaller
components´ densities are low. This means that the two smaller components are not
very compact and their structures are not very robust. The diameter of the components,
however, show that all three components display “small world” properties.
Component
No. of actors
Density
Diameter
1
25
0.2292
5
2
18
0.0948
4
3
15
0.1000
2
Table 8: Main co-authorship component characteristics for 2004.
41
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Affiliation key:
University of Maastricht, The Netherlands
University of Nijmegen, The Netherlands
University of Antwerp, Belgium
University of California, USA
Eindhoven University of Technology, The Netherlands
Portland State University, USA
Free University, Germany
Figure 20: Second component for 2004 with affiliation.
Figure 20 above displays the second component for 2004 with affiliation. It
demonstrates geographical clustering around the University of Maastricht, The
Netherlands, Eindhoven University of Technology, The Netherlands and the University
of Antwerp, Belgium, respectively. Three individual researchers from the University of
Maastricht stand out as they each are disconnected from any other researchers from that
institution.
42
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Track key:
Service Marketing
Consumer Behaviour
Relationship Marketing
Innovation and New Product Development
New Technologies and E-Marketing
Figure 21: Second component for 2004 according to track.
The second component for 2004 according to track (Figure 21 above), illustrates trackbased clustering around Bloemer and especially Wetzels. De Ruyter is the most active
node as this specific author has collaborated on four different research papers within
three different tracks.
43
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Affiliation key:
University of Nottingham, England
University of Glasgow, Scotland
University of East Anglia, England
Open University of Hong Kong
University of Strathclyde, Scotland
Cardiff Business School, Wales
Figure 22: Third component for 2004 with affiliation.
Figure 22 above shows that the third component for 2004 with affiliation exhibits
geographical clustering around the University of Nottingham, England and the
University og Glasgow, Scotland, respectively.
44
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Track key:
International and Cross-Cultural Marketing
Modeling and Forecasting
Relationship Marketing
Business-to-Business Marketing and Networks
New Technologies and E-Marketing
Consumer Behaviour
Marketing Strategy and Leadership
Figure 23: Third component for 2004 according to track.
The third component for 2004 acccording to track is presented in Figure 23 above. It
reveals no track-based clustering. It does, however, identify Ennew as the most
productive node with four different research papers written in four different tracks.
5.6 EMAC Conference 2007
In the section at hand the more thorough analysis of the EMAC conference of 2007 will
be presented.
45
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
5.6.1 Affiliation and Track
Affiliation key:
University of Mannheim, Germany
University of Maastricht, The Netherlands
Eindhoven University of Technology, The Netherlands
New University of Lisbon, Portugal
Instituto Superior de Ciencias do Trabalho e da Empresa, Portugal
Nijmegen School of Management, The Netherlands
University of Innsbruck, Germany
University of Antwerp, Belgium
No affiliation available
Figure 24: Main component for 2007 with affiliation.
The main component for 2007 according to affiliation is displayed in Figure 24 above.
It shows a large group of researchers to be affiliated with the University of Mannheim
in Germany. In addition to this, two smaller groups of researchers are affiliated with
two Dutch universities, namely the University of Maastricht and Eindhoven University
of Technology, respectively. This points to some geographical clustering within the
network. Additional analysis of the main component for 2007 revealed that although
some of the researchers are affiliated with the same university, they have co-authored
46
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
scientific research papers within a number of different tracks in the field of marketing.
This is illustrated in Figure 25 below. Stokburger-Sauer and Bauer have collaborated on
two different research papers within two different tracks, the same has Van Birgelen and
Wetzels. Wetzels, however, has co-authored two different research papers within the
same track, pointing to minor track-based clustering.
Track key:
New Technologies and E-Marketing
Marketing of Public and Non-Proft Organisations
Service Marketing
International and Cross-Cultural Marketing
Marketing Strategy and Leadership
Business-to-Business Marketing and Networks
Tourism Marketing
Social Responsibility, Ethics and Consumer Protection
Advertising, Promotion and Marketing Communication
Figure 25: Main component for 2007 according to track.
5.6.2 Central Actors
The most active researcher in the main component for 2007 is, as can be seen in Table 9
below, Wetzels. He is the best connected researcher in the network as reflected in his
47
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
actor degree centrality score and repeated in his actor closeness centrality score. The
actor betweenness centrality measure reveals that Wetzels also has the highest number
of indirect connections through his direct links in the network. These findings are in
accordance with the diagram in Figure 8, which also identifies Wetzels as the main
actor in 2007.
Actor
Actor
Actor
Actor
Neighbourhood degree closeness betweenness
size
centrality centrality
centrality
Wetzels
11
11
54.762
Bauer
8
9
42.593
Lageslu
7
7
41.818
Table 9: Actor centrality measures for the main component of 2007.
181.500
110.500
62.000
Table 10 below shows additional centrality measures for the main component of 2007.
Oddly enough, all the eigenvector centrality scores for 2007 were negative. Everett
(2001) found a similar result and gave the explanation that the negative results were due
to the network containing disconnected components. He used the absolute value of the
scores instead. Following Everett (2001), the absolute value of the eigenvector
centrality score will be used in interpreting the findings. This then locates Wetzels as
the actor with the most well connected connections. He also holds the highest flow
betweenness centrality score and it would therefore reduce the flow in the network if he
was to leave the network. The reach centrality measure points to a less dense network as
the most central actor, Wetzels, can only reach 74 % percent of the component in two
steps. Finally, the concept of structural holes reveals Wetzels to be the researcher who
links researchers that would otherwise be disconnected. These findings can also be seen
in Figure 8.
Actor
Flow
Reach centrality
Eigenvector betweenness
Structural
centrality
centrality
two-step
three-step
hole
Wetzels
- 73.648
370.000
0.74
Bauer
- 27.323
226.000
0.43
Lageslu
- 50.084
128.000
0.61
Table 10: Additional centrality measures for the main component of 2007.
0.96
0.87
0.74
9.000
6.556
5.000
48
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
5.6.3 Additional Component Analysis
The following results refer to the cumulative network for 2007. Table 11 below reveals
the main component together with the second and third largest components in the
cumulative network for 2007. The density of the components are relatively small
indicating a sparsely connected network. The diameter of the components, however,
show that all three components display “small world” properties.
Component
No. of actors
Density
Diameter
1
24
0.0797
4
2
19
0.0965
5
3
14
0.1264
3
Table 11: Main co-authorship component characteristics for 2007.
Affiliation key:
Delft University of Technology, The Netherlands
University of Wollongong, Australia
Erasmus University, The Netherlands
Wirtschaftsuniversität, Wien, Austria
McMaster University, Canada
Graz University, Austria
49
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Concordia University, Montreal, Canada
Bishop´s University, Canada
No affiliation available
Figure 26: Second component for 2007 with affiliation.
Figure 26 above illustrates the second component for 2007 according to affiliation. It
reveals geographical clustering around Delft University of Technology, The
Netherlands and the University of Wollongong, Australia.
Track key:
Innovation and New Product Development
Advertising, Promotion and Marketing Communication
Product and Brand Management
Marketing Research and Research Methodology
Marketing of Public and Non-Profit Organisations
Figure 27: Second component for 2007 according to track.
The second component for 2007 according to track is displayed in Figure 27 above. The
component exhibits relatively strong track-based clustering with six papers being co-
50
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
authored within the same track. Minor track-based clustering also appears around El
Houssi.
Affiliation key:
University of Groningen, The Netherlands
Erasmus University, The Netherlands
Waikato Management School, New Zealand
TNO –Netherlands Organisation for Applied Scientific Research, The Netherlands
VU University of Amsterdam, The Netherlands
Tilburg University, The Netherlands
Delft University of Technology, The Netherlands
No affiliation available
Figure 28: Third component for 2007 with affiliation.
Figure 28 above presents the third component for 2007 according to affiliation. The
figure demonstrates geographical clustering around the University of Groningen, The
Netherlands.
51
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Track key:
New Technologies and E-Marketing
Marketing Research and Research Methodology
Relationship Marketing
Consumer Behaviour
Innovation and New Product Development
International and Cross-Cultural Marketing
Figure 29: Third component for 2007 according to track.
The third component for 2007 according to track is presented in Figure 29 above. The
component only displays minor track-based clustering around Huizingh. The most
active node in the component is Bijmolt, who has co-authored six research papers in six
different tracks.
5.7 EMAC Conference 2010
In the current section the results of the more detailed analysis of the EMAC conference
of 2010 will be presented.
52
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
5.7.1 Affiliation and Track
Affiliation key:
University of Hamburg, Germany
University of Cologne, Germany
University of Innsbruck, Germany
Erasmus University Rotterdam, The Netherlands
Waikato Management School, New Zealand
Groningen University, The Netherlands
University of Munich, Germany
University of St. Gallen, Switzerland
Bauhaus University of Weimar, Germany
Figure 30: Main component for 2010 with affiliation.
Figure 30 above shows the main component for 2010 according to affiliation. The
diagram reveals three large groups of researchers to be affiliated with the University of
Hamburg, the University of Cologne and the University of Innsbruck, respectively.
Additionally, a small group of researchers from the University of Hamburg is
disconnected from the larger group of researchers from this specific university but
connected to the group through Völckner, who herself is affiliated with the University
of Cologne. These findings indicate strong geographical clustering within the network.
53
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Further analysis of the main component for 2010 showed a moderate tendency towards
track-based clustering grouped around Sattler and Völckner. This is demonstrated in
Figure 31 below. Sattler has co-authored two research papers in each of two different
tracks. Völckner, in particular, has been very active collaborating in three different
tracks with a total of six scientific research papers.
Track key:
Advertising, Promotion and Marketing Communication
Consumer Behaviour
Innovation and New Product Development
Product and Brand Management
Service Marketing
International and Cross-Cultural Marketing
Marketing Research and Research Methodology
Marketing Theory
Marketing of Public and Non-Profit Organisations
Figure 31: Main component for 2010 according to track.
54
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
5.7.2 Central Actors
According to Table 12 below, Völckner is the most active researcher in the main
component for 2010. The actor degree centrality measure score of Völckner reflects a
high degree of direct contact with many other actors in the network. She also serves as
the actor who is closest to other actors in the network, which is demonstrated by the
relatively high actor closeness centrality score. The actor betweenness centrality
measure supports the results of the two former as it identifies Völckner as the leading
researcher in the network. These findings are reflected in the diagram in Figure 11, in
which it can be seen, that Völckner holds a central position in the network.
Actor
Actor
Actor
Actor
Neighbourhood degree closeness betweenness
size
centrality centrality
centrality
Völckner
16
17
47.368
Sattler
8
10
37.500
Matzler
7
7
24.490
Table 12: Actor centrality measures for the main component of 2010.
447.000
158.500
99.000
Table 13 below displays additional centrality measures for the main component of 2010.
Not surprisingly, Völckner has the highest eigenvector centrality score as she is
connected to many other actors who are also well connected (see Figure 11). Matzler´s
eigenvector centrality score, however, is low. The reason for this can be seen in Figure
11. It shows Matzler to be connected to a group of researchers, who are only connected
to the rest of the component through one researcher. The flow betweenness centrality
measure supports what can be seen in Figure 11, namely that if Völckner decides to
leave the network it would leave it scattered. The reach centrality measure indicates a
less dense network as the most central actor can only reach 72 % percent of the
component in two steps. The structural hole measure confirms what the other measures
have demonstrated, that Völckner occupy a hub position in the network.
55
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Actor
Flow
Reach centrality
Eigenvector betweenness
Structural
centrality
centrality
two-step
three-step
hole
Völckner
79.036
913.000
0.72
Sattler
58.989
360.333
0.64
Matzler
0.551
203.000
0.25
Table 13: Additional centrality measures for the main component of 2010.
0.81
0.75
0.31
13.912
6.100
4.429
5.7.3 Additional Component Analysis
The following results refer to the cumulative network for 2010. The main component
together with the second and third largest components in the cumulative network for
2010 is presented in Table 14 below. The density of the components are relatively low
indicating a sparsely connected network. This is also reflected in the diameter of the
main component as it is above the threshold value for exhibiting “small world”
properties. The diameters for the second and third largest components for 2010 do
indeed, however, diplay “small world” properties.
Component
No. of actors
Density
Diameter
1
37
0.0541
6
2
24
0.0833
3
3
15
0.0952
2
Table 14: Main co-authorship component characteristics for 2010.
56
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Affiliation key:
Ghent University, Belgium
Catholic University of Leuven, Belgium
University of Vienna, Austria
Vlerick Leuven Gent Management School, Belgium
University of Bern, Switzerland
Richard Ivey School of Business, Canada
Figure 32: Second component for 2010 with affiliation.
Component number two according to affiliation is presented in Figure 32 above. A
relatively strong geographical clustering around Ghent University, Belgium becomes
evident when examining the diagram. The figure also reveals minor geographical
clustering around the Catholic University of Leuven, Belgium.
57
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Track key:
Consumer Behaviour
Innovation and New Product Development
Product and Brand Management
Marketing Research and Research Methodology
Retailing, Channel Management and Logistics
Advertising, Promotion and Marketing Communication
International and Cross-Cultural Marketing
Figure 33: Second component for 2010 according to track.
Figure 33 above illustrates component number two according to track for 2010. It
reveals strong track-based clustering around Vermeir, in particular. This specific actor is
also the most productive member of the component with seven different research papers
written in three different tracks.
58
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Affiliation key:
University of Murcia, Spian
Catholic University of Leuven, Belgium
BI – Norwegian School of Management, Norway
Autonomous University Madrid, Spain
Koc University, Turkey
Figure 34: Third component for 2010 with affiliation.
Component number three for 2010 according to affiliation is displayed above in Figure
34. It presents geographical clustering around the University of Murcia, Spain.
Additionally, minor clustering appears around the Catholic University of Leuven,
Belgium and the Norwegian School of Management, Norway.
59
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Track key:
Consumer Behaviour
Advertising, Promotion and Marketing Communication
New Technologies and E- Marketing
Social Responsibility, Ethics and Consumer Protection
Figure 35: Third component for 2010 according to track.
The final component to be displayed is the third component for 2010 according to track
(Figure 35). It shows relatively strong track-based clustering around Warlop within the
track Consumer Behaviour. Furthermore, clustering appears in the track Advertising,
Promotion and Marketing Communication. The most active node in the network is
Ruizsa, who has collaborated on four different research papers in three different tracks.
6. Discussion of Results
The following sections will discuss the findings presented above. The first section will
discuss the development of the EMAC network from 200 to 2010 and visualise findings
from Table 2. The four succeeding sections will then discuss the findings of the analysis
of the years 2001, 2004, 2007 and 2010 individually.
60
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
6.1 Development of the EMAC Network 2000-2010
In reviewing the literature, several studies revealed research communities that were not
fully connected as the networks were comprised of a number of subgroups not
connected to each other by paths between authors (Vidgen et al., 2007; Henneberg et al.,
2009). These results are in line with the findings of the study at hand, which also
observed a number of components without links between them - a result that applied to
both the cumulative network and the individual years. This can, in part, be ascribed to
researchers collaborating on scientific research papers by preferential attachment, close
geographical or cultural proximity, or similar field of interest in research.
For the cumulative network (i.e. 2000-2010) more than half of all actors are in the main
component. This corroborates the findings of previous studies on co-authorship
(Newman, 2001a; Liu et al., 2005; Henneberg et al. 2009). This result points to a
tendency for some authors to work across different tracks and topics in the field of
marketing.
1000
No. of actors in network
900
800
700
600
500
400
300
200
100
0
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
Year
Figure 36: The development in the number of actors in the EMAC network 2000-2010.
The development in the number of actors in the EMAC network 2000-2010 is visualised
in Figure 36 above. It reveals a general increase in the number of researchers who have
participated in the EMAC conferences with one or more co-authored research papers.
This finding is consistent with the results of Laband and Tollison (2000) and Moody
(2004) who observed an increase in co-authorship in a number of scientific disciplines.
A possible explanation for the increase in researchers co-authoring in the EMAC
61
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
community could be an increased degree of specialisation. When examining the main
components according to track in three of the four years which are the focus of the
current study, a tendency towards specialisation becomes evident. The figures reveal
that the central authors tend to change collaboration partners according to the topic of
the research paper. Interviews with central actors in the network would provide
information on the degree to which researchers in the EMAC community tend to
specialise.
A strong relationship between the measures density and diameter and the concept of
“small worlds” has been reported in the literature (Newman, 2001b; Barabási et al.,
2002; Morlacchi et al., 2005; Vidgen et al., 2007). The findings of the study at hand
lends support to this relationship. Figure 37 below illustrates the development in the
density of the main components in the period 2000 to 2010. It shows that the density is
generally rather low with the exception of the year 2002 and to some extent 2004. This
points to the networks not exhibiting “small world” characteristics. These generally low
scores indicate that the networks of the individual EMAC conferences are somewhat
sparsely connected. A closer inspection of Table 2 reveals the reason behind this result.
The individual years which have the highest number of participating actors display the
lowest density score. This result is in agreement with the findings of Vidgen et al.
(2007) and Henneberg et al. (2009), which demonstrated an asymptotic convergence of
density towards 0.0 as the number of actors in the network increased. The
comparatively low scores in the current network means that information does not travel
very fast within the network, i.e. new research methods, theories and concepts are not
easily diffused in the network of the individual years.
62
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
0,5
0,45
0,4
Density
0,35
0,3
0,25
0,2
0,15
0,1
0,05
0
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
Year
Figure 37: The development in density of the main components 2000-2010.
The development in the diameter of the main components in the period 2000 to 2010 is
visualised in Figure 38 below. Despite the relatively low density scores, the majority of
the individual years, with the exception of 2009 and 2010, display “small world”
properties as their diameters are 5 or lower. Figures 1 to 11 demonstrate the majority of
years to have networks which are grouped around central researchers who collaborate
extensively within the network and play a crucial role in linking various parts of the
network, i.e. they display “small world” properties. These findings are in accordance
with results presented by Newman (2001b), Barabási et al. (2002) and Morlacchi et al.
(2005) as they found the typical distance between researchers within networks to be
small. However, the findings of the study at hand do not support previous research on
co-authorship in a conference setting. On the contrary, Vidgen et al. (2007) and
Henneberg et al. (2009) found that the main components of the collaboration networks
under their investigation displayed “large world” properties. These contradicting results
may be explained by the fact that these two studies were based on cumulative data
whereas the current study focuses on the main components of the individual years.
The fact that the majority of the individual years exhibit “small world” characteristics
points to a relatively easy diffusion of new theories, concepts and research methods
within the network.
63
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
7
6
Diameter
5
4
3
2
1
0
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
Year
Figure 38: The development in diameter of the main components 2000-2010.
Prior studies have noted the presence of so-called “invisible colleges” within social
networks (Crane, 1972; Eaton et al., 1999; Morlacchi et al., 2005; Henneberg et al.,
2009). The study at hand produced results which corroborate the findings of previous
work in this field. “Invisible colleges” formed by highly productive and central actors
take up hub positions in the main components of the current study. These researchers
hold key roles in disseminating knowledge throughout the network as they reoccur as
central players in a number of years. An example of an “invisible college” within the
EMAC community is a group of ten researchers (Anttila, Matear, Hooley, Theoharakis,
Greenley, Tuominen, Hyvönen, Möller, Rajala and Kasper), who are all members of the
compact main component in the year 2004. These actors reoccur in 2006 where they
form the core of the main component for that year. This finding indicates continuous
collaboration over time, hence supporting the notion of “invisible colleges” within the
EMAC community.
6.2 EMAC Conference 2001
The present study was designed to determine the structure of EMAC´s collaboration
network and the way in which researchers in the EMAC community work together to
generate knowledge. The most interesting result from the year 2001 was that the main
component displayed strong geographical clustering around in particular, Ghent
64
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
University, Belgium and Maastricht University, The Netherlands. A possible
explanation for this might be preferential attachment as some actors might prefer to
collaborate with actors of close geographical and cultural proximity to themselves. This
could be due to the fact that it might be easier to work with colleagues from the same
workplace that speak the same language. Another possible explanation could be that this
group of researchers interact socially and have developed friendships which in turn has
developed into professional working relationships. The results of the current study
supports the findings of previous research which found a tendency for geographical
groupings within networks (Liu et al., 2005; Morlacchi et al., 2005).
Another interesting result is the way in which the actors in the main component are
divided according to track. The authors in this specific main component work across a
diverse set of tracks. However, the network does not display track-based clustering as
each individual actor has only co-authored one research paper in each track. A possible
explanation for the lack of track-based clustering could be the fact that the main
component for 2001 is extremely small. A higher number of actors in the main
component would lead to a greater possibility of actors working across more than one
track.
It is interesting to note that the most well connected actor in the main component holds
the highest score on all the centrality measures. This identifies De Wulf as the most
prominent researcher in the main component for 2001 as this researcher takes up a
strategically important position within the network. It suggests that De Wulf is the most
popular collaboration partner and other actors will then more likely acknowledge the
central location in the network of this researcher as a leading channel of relational
information. With a high score on all the centrality measures De Wulf is very likely to
be the researcher to hear of new research methods, theories, terms and concepts first as
this actor is well connected. This actor can thereby influence the flow of information in
the network. Frequently published actors like De Wulf seem to be generators of stucture
within the network. The collaborations of these actors create the links that shed light on
the institutional, intellectual and social structure of the EMAC community.
An additional observation from the diagrams in Figure 12 and Figure 13 is that the more
productive authors tend to collaborate more and to collaborate across tracks. The three
65
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
central actors in the main component have all worked on more than one track. This
observation could lend support to the tendency towards specialisation as it appears that
the central actors bring in specialists when working on different tracks. However, more
research on this topic needs to be undertaken before the association between the way in
which productive actors collaborate across tracks and specialisation is more clearly
understood.
Classical network models work under the assumption of complete randomness, i.e. the
actors are connected to each other independently of the number of links they already
have (Watts, 1999). However, the findings of the study at hand do not support this
assumption. The main component for 2001 revealed itself to be highly clustered around
a few individual researchers and the network is therefore not characterised by a random
set of connections. The central actors of the network serve as centers from which
knowledge (and possibly also research strategy) is pushed out. The structure of the main
component of 2001 does therefore not reflect a random process but rather an intentional
form of research collaboration.
Additional analysis of the second and third largest component in the cumulative
network for 2001 provided further support for the suggestion that members of the
EMAC community tend to work with authors of close geographical and cultural
proximity. The second component revealed relatively strong clustering while
component three exhibited very strong clustering with all members being affiliated with
the same institution. The findings of the additional component analysis also supply
support for the results of the analysis of the main component with regards to trackbased clustering. The second component demonstrated only minor clustering while the
third component displayed none. These findings indicate a tendency for choosing
collaboration partners based on geographical and cultural proximity for the network of
2001.
6.3 EMAC Conference 2004
In contrast to the findings for the year 2001, the results for the year 2004 showed very
limited geographical clustering in the main component. Helsinki School of Economics
in Finland is the most well represented institution with four affiliated researchers. This
66
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
finding is somewhat surprising when compared to the results from 2001. A possible
explanation for this, however, emerges when inspecting Figure 19, which displays the
main component according to track. It reveals very strong track-based clustering as all
members of the main component have collaborated within one single track. This
indicates that members of the main component choose collaboration partners based on
their field of interest, i.e. the track. Caution should be taken with this result, however.
As the main component is comprised mainly of authors collaborating on just one
research paper it is difficult to make general statements about the behaviour of this
group of actors based on the findings from this component.
At first glance the results presented for the centrality measures for the three most central
actors are surprising as they are identical. Closer inspection of Figure 19 and the
conference proceeding for 2004, however, shows that these actors have identical
collaboration partners and have co-authored the same research papers. In doing so they
hold identical positions in the network. More researchers would turn out to hold
equivalent positions in the network if the list of central actors was expanded. This is due
to the fact that the majority of the actors in the main component for 2004 have coauthored one research paper together. Further research should be done to investigate the
motivation behind this large number of actors collaborating on this specific research
paper. This could be done in the form of interviews.
Figure 18 and 19 do not provide support for the connection between author productivity
and collaboration. As the main component of 2004 has proven to be quite extraordinary,
it will be relevant at this point to include the second and third component to discover if
the cumulative network provided support for the observations made for the main
component. Figures 20 to 23 supply support for the findings from the year 2001 which
found that the more productive nodes tend to collaborate more and that they tend to
collaborate across tracks. These results cement the fact that the results obtained from the
main component of 2004 are quite exceptional.
Like the year 2001, 2004 does not support the assumption of complete randomness. In
displaying such strong track-based clustering it must be assumed that collaboration
partners are chosen on the basis of their field of interest and knowledge within a field
and not on a random basis.
67
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
In contrast to the findings of the main component for 2004, the second and third
component revealed some geographical clustering. This result corroborates the findings
of both the main component and the second and third component of 2001. The
additional component analysis for 2004 provided dissimilar results with regards to
track-based clustering. The main component displayed very strong track-based
clustering while the second component showed some clustering, however, not nearly as
strong as the main component. The third component contradict the main component as
it exhibited no clustering. These findings indicate a tendency for the members of the
second component to choose collaboration partners on the basis of both close
geographical and cultural proximity and field of interest. In the third component close
geographical proximity appears to be the main reason for collaboration partner
selection.
6.4 EMAC Conference 2007
The most interesting finding from the year 2007 was that the main component exhibited
some geographical clustering, especially around the University of Mannheim, Germany.
Two other geographical clusters are comprised of researchers from the University of
Maastricht, The Netherlands and Eindhoven University of Technology, The
Netherlands, respectively. This result may be explained by a preference for working
with people of close cultural and geographical proximity as described above in the
section on the EMAC conference of 2001. Even though geographical clustering is
present in the main component only four out of the twelve research papers in the
component are co-authored by actors from the same institution. An additional
interesting finding is the manner in which the researchers in the main component are
separated according to track. Despite the fact that the main component is relatively large
and the possibility of tracks overlapping therefore would be larger, it only showed
minor track-based clustering. These results point towards a tendency for members of the
main component to choose collaboration partners based mainly on close
geographical/cultural proximity and to a minor degree according to field of interest.
As in the main component of 2001, the most well connected node in the main
component of 2007 has the highest score on all the centrality measures. Wetzels is in
contact with or adjacent to many other actors in the network and can then be
68
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
characterised as the most important actor in the network as he occupies a strategically
important position. In holding a central position in the network Wetzels is closer to
other members of the network than any other actor. This means that he can access
information rather quickly and that he will serve as a reference point for other nodes in
the network. Like De Wulf´s links in 2001, the connections of Wetzels can clarify the
structure of the EMAC community.
The main component of 2007 supports the observation concerning author productivity
and collaboration made for the year 2001. Figure 24 and 25 revealed that the most
productive actors tend to collaborate more and to collaborate across tracks. Again all
three central actors in the main component have worked on numerous tracks. This lends
support to the suggestion that the EMAC community to some degree is comprised by
specialists. However, as mentioned above, further research on this should be conducted
before making a general assumption.
In accordance with the findings of 2001 and 2004, the main component of 2007 does
not corroborate the assumption of complete randomness. Collaboration partners seem to
be selected on the basis of both close geographical/cultural proximity and according to
field of interest, as mentioned above.
Further investigation of the second and third largest components in the cumulative
network provided support for the original results from the main component, i.e. they
also displayed geographical clustering. Both the second and the third component
exhibited some clustering lending support to the suggestion that actors in the EMAC
community are inclined to work with researchers of close geographical and cultural
proximity. The additional analysis of the second component revealed relatively strong
track-based clustering contradicting the findings from the main component. This
indicates that the nodes in this specific component choose collaboration partners based
on both geographical proximity and field of interest. The third component supports the
initial findings of the main component as it displays minor track-based clustering. As
with the main component, the members of the third component are inclined to select
collaboration partners mainly on the basis of close geographical and cultural proximity
and to a lesser extent because of field of interest.
69
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
6.5 EMAC Conference 2010
The study at hand found the main component of 2010 to display strong geographical
clustering around the universities of Hamburg, Cologne and Innsbruck, respectively.
This result can, as mentioned above, be explained by a preference of the members of the
main component to collaborate with authors of close geographical and cultural
proximity. Another interesting finding is the way in which the nodes in the main
component are divided according to track. The actors in the main component work in
many different tracks but some degree of track-based clustering still emerges around
two actors. These results point to members of the main component choosing
collaboration partners on the basis of both close geographical and cultural proximity
and field of interest.
Like the years 2001 and 2007, the most well connected actor in the main component of
2010 has the highest score on all the centrality measures. This identifies Völckner as a
significant actor in the main component for 2010 as she occupies a hub position in the
network. As with the central actors in main components discussed above, Völckner is a
popular collaboration partner and will therefore more likely hear of new research
methods, terms and concepts before other actors who are not so well connected. She
will then be a mediator for information in the network.
The main component for 2010 confirms the observations made for the main components
of 2001 and 2007 and for the additional components of 2004 suggesting a connection
between author productivity and collaboration (see Figure 30 and 31). Again the three
most active nodes have all collaborated on more than one track. This discovery
corroborates the suggestion made above that the authors in these components tend
change collaboration partners when they work within different fields in marketing.
Again caution should be taken in making a general assumption before additional
research on this result has been made.
The results of the analysis of the main component of 2010 provided further support for
the observations made regarding randomness in the other three years. The members of
the main component of 2010 appear to select collaboration partners based on either
close geographical and cultural proximity or field of interest and not at a random basis.
70
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Additional analysis of the second and third largest components in the cumulative
network of 2010 revealed the same patterns with regard to geographical clustering as in
the main component, i.e. strong clustering. However, the two additional components
exhibited very strong track-based clustering. These results indicate a tendency for
members of the second and third component in the year 2010 to choose collaboration
partners according to field of interest rather than close geographical proximity.
7. Limitations
A number of limitations exist for the study at hand. First of all, the current study
considers only one aspect of research collaboration, namely co-authorship in a
conference setting. By omitting books and journal articles outside the EMAC
conference setting significant contributions to knowledge generation are not considered.
Second, single authored research papers are not included in the study and this offers a
drawback to the methodology adopted in the present study as researchers publishing
alone also add to the generation of knowledge within a field. Third, a possibility that
two authors co-authored a paper before 2000 exists, but in the present data set they
appear as disconnected because the focus of this study is only on 2000-2010. This poses
a methodological disadvantage to the study but is a consequence of the limited time
frame studied.
8. Further Research
The present study provides the basis for further studies of the way in which members of
the EMAC community collaborate on scientific research papers. It would be relevant to
perform a series of interviews with actors of the EMAC network in order to make more
accurate assumptions about collaboration in the network. An interesting topic in the
interviews would be the year 2004 as it would be relevant to establish the motivation
behind the research paper that formed the basis of the main component of 2004. As this
year proved to be quite exceptional it would be interesting to find out why so many
authors collaborated on this specific research paper. The interviews should also be
designed to give information on the degree to which researchers in this specific network
tend to specialise. This will provide a deeper insight into the way in which actors
choose collaboration partners. An additional yield from the interviews should be more
information regarding the more productive researchers in the community. Why do they
71
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
collaborate more? And why do they appear to collaborate across more tracks than less
productive actors?
For future research purposes, it would also be interesting to explore how researchers
initially join the EMAC community. Is it through association with frequently published
authors? Such as when Ph.D. students publish with their dissertation advisor. Or is it
because they are affiliated with institutions which are productive in EMAC context and
thus participate with numerous research papers in every annual conference?
In order to acquire more information on how knowledge is generated and shared among
this group of researchers it could be relevant to extent the analysis to include citation
analysis. This will add to the understanding of the structure of the network and provide
a more complete picture of the network as a whole.
The strength of the ties between the actors in the EMAC community, i.e. the number of
research papers any two actors have produced together, has only been mentioned briefly
in the study at hand. Additional research on the influence of tie strength on the
collaboration structure of the network could be interesting.
9. Conclusion
The present study has demonstrated the application of Social Network Analysis to coauthorship in a conference setting. The objective of the study was to analyse the
structure of the collaboration network of the EMAC community and to unveil patterns
and regularities in the choice of collaboration partners of the members of this specific
community over time.
The structure of the EMAC scientific research community is characterised by a number
of highly visible authors who dominate the individual years. These researchers occupy
central positions in the network from which knowledge (and possibly also research
strategy) is pushed out. In holding key positions in the network these actors will
automatically become leading channels of relational information and will naturally
become popular collaboration partners. As the central nodes in the network are more
72
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
likely to hear of new research methods, terms, concepts and theories first they serve as
reference points for other actors in the network.
A very interesting finding to emerge from the study was that the central and more
productive actors in the EMAC community tend to collaborate more and that they tend
to collaborate in a more diverse set of tracks than other less active authors. This could,
as mentioned earlier, be an interesting topic for further research.
The purpose of the current study was also to determine which factors influence the
choice of collaboration partners in the EMAC research community. The study at hand
provided varying results for the three components in the four years under investigation.
The primary factor of influence when choosing collaboration partners was found to be
close geographical and cultural proximity as indicated in the majority of the
components in the analysis. This points to a tendency for members of the EMAC
community to be organised around institutions. In a number of the components in the
study at hand the selection criteria was a combination of close geographical/cultural
proximity and field of interest as both factors appeared to equally influence the choice
of collaboration partners. In other components researchers relied mainly on the field of
interest when selecting collaboration partners. This was especially true for the main
component of 2004 which proved to be quite exceptional. As the results generated by
this component were exraordinary no general conclusions about the network should be
drawn from this specific component.
In accordance with previous research the present study has shown a general increase in
the number of authors who have participated in the EMAC conferences with coauthored research papers. As suggested this result could be an indication of increased
specialisation in the network. This is further supported by the fact that central actors
tend to change collaboration partners according to the topic of the research paper.
Social network analysis of co-author networks gives an interesting insight into the
sociology of scientific workers. The findings of the study at hand adds substantially to
the understanding of the way in which scholars in the European Marketing Academy
work together to produce knowledge. It has showed that members of this research
73
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
community have certain criteria for selecting collaboration partners instead of random
selection.
74
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
References
A
Acedo, Fransisco José, Barroso, Carmen, Casanueva, Cristóbal and Galán, José Luis
(2006): Co-authorship in Management and Organizational Studies: An Empirical
and Network Analysis. Journal of Management Studies, 43 (5), pp. 957-983.
B
Barabási, A. L., Jeong, H., Néda Z., Ravasz, E., Schubert, A. and Vicsek, T. (2002):
Evolution of the Social Network of Scientific Collaborations. Physic A: Statistical
Mechanics and its Applications, 311 (3-4), pp. 590-614.
Borgatti, S. P., Everett, M. G. and Freeman, L. C. (1999): UCINET 6.0 Version 1.00.
Natick: Analytic Technologies.
Bourdieu, Pierre (2005): Viden om viden og refleksivitet. Copenhagen, Hans Reitzels
Forlag.
Burt, Ronald S. (1995): Structural Holes: The Social Structure of Competition.
Cambridge, Massachusetts and London, England: Harvard University Press.
C
Carter, Craig R., Leuschner, Rudolf and Rogers, Dale S. (2007): A Social Network
Analysis of the Journal of Supply Chain Management: Knowledge Generation,
Knowledge Diffusion and Thought Leadership. The Journal of Supply Chain
Management, 43 (2), pp. 15-28.
Crane, Diana (1972): Invisible Colleges – Diffusion of Knowledge in Scientific
Communities. Chicago & London. The Univeristy of Chicago Press.
Cross, Rob, Parker, Andrew, Prusak, Laurence and Borgatti, Stephen (2001): Knowing
What We Know: Supporting Knowledge Creation and Sharing in Social
Networks. Organisational Dynamics, 30 (2), pp. 100-120.
Cross, Rob, Borgatti, Stephen and Parker, Andrew,
(2002): Making the Invisible
Visible: Using Social Network Analysis to Support Strategic Collaboration.
California Management Review, 44 pp. 25-46.
D
Dekimpe, Marnik G. (2010): From the editor of IJRM. The EMAC Chronicle, 7, May
2010, pp. 9.
75
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
E
Eaton, John P., Ward, James C., Kumar, Ajith and Reingen, Peter H. (1999): Structural
Analysis of Co-author Relationships and Author Productivity in Selected Outlets
for Consumer Behavior Research. Journal of Consumer Psychology, 8 (1), pp. 3959.
EMAC Conference Proceeding 2000: Marketing in the New Millenium. EMAC 29th
Conference. Conference Proceedings. Rotterdam, The Netherlands. Pp. 1-190.
EMAC Conference Proceeding 2001: Rethinking European Marketing. EMAC 30th
Conference. Conference Proceedings. Bergen, Norway. Pp. 1-149.
EMAC Conference Proceeding 2002: Marketing in a Changing World – Scope,
Opportunities and Challenges. EMAC 31st Conference. Conference Proceedings.
Braga, Portugal. Pp. 1-213.
EMAC Conference Proceeding 2003: Marketing: Responsible and Relevant? EMAC
32nd Conference. Conference Proceedings. Glasgow, Scotland. Retrived from CD.
EMAC Conference Proceeding 2004: Worldwide Marketing? EMAC 33rd Conference.
Conference Proceedings. Murcia, Spain. Pp. 1-225.
EMAC Conference Proceeding 2005: Rejuvenating Marketing: Contamination.
Innovation. Integration. EMAC 34th Conference. Conference Proceedings. Milan,
Italy. Retrieved from CD.
EMAC Conference Proceeding 2006: Sustainable Marketing Leadership – A Synthesis
of Polymorphous Axioms, Strategies and Tactics. EMAC 35th Conference.
Conference Proceedings. Athens, Greece. Pp. 1-256.
EMAC Conference Proceeding 2007: Flexible Marketing in an Unpredictable World.
EMAC 36th Conference. Conference Proceedings. Reykjavik, Iceland. Pp. 1-463.
EMAC Conference Proceeding 2008: Marketing Landscapes: A Pause for Thought.
EMAC 37th Conference. Conference Proceedings. Brighton, United Kingdom.
Retrieved online from http://www.emac-online.org/r/default.asp?iId=IIFKL.
EMAC Conference Proceeding 2009: Marketing and the Core Disciplines:
Rediscovering References? EMAC 38th Conference. Conference Proceedings.
Nantes, France. Pp. 1-324.
EMAC Conference Proceeding 2010: The Six Senses – The Essentials of Marketing.
EMAC 39th Conference. Conference Proceedings. Copenhagen, Denmark. Pp. 1351
76
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Everett, Martin (2001), personal communication, March 8, 2001. Cited in: Van Stelle,
Jennifer (2001): Centrality and Influence i the Semiconductor Industry
Association: An Exploratory Study. Presented at the International Sunbelt Social
Network Conference, New Orleans, LA in February 2002.
G
Granovetter, Mark S. (1973): The Strength of Weak Ties. The American Journal of
Sociology, 78 (6), pp. 1360-1380.
H
Hanneman, Robert A. and Riddle, Mark (2005): Introduction to Social Network
Methods. Riverside, USA: University of California. Retrieved Juli 10th 2010 at
http://faculty.ucr.edu/~hanneman/
Henneberg, Stephan, Jiang, Zhizhong, Naudé, Peter and Ormrod, Robert (2009): The
Network Researchers´ Network: A Social Network Analysis of the IMP Group
1984-2006. The IMP Journal, 3 (1), pp. 28- 49.
K
Knoke, David and Yang Song (2008): Social Network Analysis (second edition). USA.
SAGE Publications. Series: Quantitative Applications in the Social Sciences.
L
Laband, David N. and Tollison, Robert D. (2000): Intellectual Collaboration. The
Journal of Political Economy, 108 (3), pp. 632-662.
Liu, Xiaoming, Bollen, Johan, Nelson, Michael L. and Van de Sompel, Herbert (2005):
Co-authorship in the Digital Library Community. Information Processing and
Management, 41, pp. 1-28.
Lotka, A. J. (1926): The Frequency Distribution of Scientific Productivity. Journal of
the Washington Academy of Science, 16 (12), pp. 317-323.
M
Moody, James (2004): The Structure of a Social Science Collaboration Network:
Disciplinary Cohesion from 1963 to 1999. American Sociological Review, 69 (2),
pp. 213-238.
Morlacchi, P., Wilkinson, I. F. and Young, L. C. (2005): Social Networks of
Researchers in B2B Marketing: A Case Study of the IMP Group 1984-1999.
Journal of Business-toBusiness Marketing, 12 (1), pp. 3- 34.
77
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
N
Newman, M. E. J. (2001a): The Structure of Scientific Collaboration Networks.
Proceedings of the National Academy of Science, 98 (2), pp. 404-409.
Newman, M. E. J. (2001b): Who is the Best Connected Scientist? A Study of Scientific
Coauthorship Networks. Physics Review E, 64, pp. 1-17.
V
Vidgen, Richard, Henneberg, Stephan and Naudé, Peter (2007): What Sort of
Community is the European Conference on Information Systems? A Social
Network Analysis 1993-2005. European Journal of Information Systems, 16, pp.
5-19.
W
Wasserman, Stanley and Faust, Katherine (1994): Social Network Analysis – Methods
and Applications. New York: Cambridge University Press.
Watts, Duncan J. (1999): Networks, Dynamics, and the Small-World Phenomenon. The
American Journal of Sociology, 105 (2), pp. 493-527.
Watts, Duncan J. and Strogatz, Steven H. (1998): Collective Dynamics of “SmallWorld” Networks. Nature, 393, pp. 440-442.
Wong, Veronica (2010): EMAC 2010 Membership Survey: Key Findings. The EMAC
Chronicle, 7, May 2010, pp. 5-7.
Z
Zinkhan, George M., Roth, Martin S. and Saxton Mary Jane (1992): Knowledge
Development and Scientific Status in Consumer-Behavior Research: A Social
Exchange Perspective. Journal of Consumer Research, 19, pp. 282-291.
Websites
http://www.emac-online.org/r/printPage.asp?iID=IHGMD
http://www.emac-online.org/r/printPage.asp?iID=IIKFF
78
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Appendix 1: Code Names
Name
Code name
Adams, Edward I.
Adams, Leen
Ahmed, Pervaiz K.
Ahmed, Sadrudin A.
Ali, Abdul Manan
Ali, Haider
Ali, M. Yunus
Alvarez, Begoña
Alvarez, M. Teresa
Alvarez
Andersen, AnneMette Sonne
Andersen, Otto
Andersen, Poul
Houman
Anderson, Helén
Anderson, Susan
Andrews, Jonlee
Andrews, Lynda
Anthony, Christina
Anthony, Janine
Antonella, Carú
Antonella, Cugini
Armario, Enrique
Martin
Armario, Julia
Martin
Arnold, Armin
Arnold, Mark J.
Backhaus, Christof
Backhaus, Klaus
Baker, Ellen
Baker, Michael
Baker, Susan
Baker, William
Bal, Charles
Bal, Rene
Banerjee,
Madhumita
Banerjee,
Subhabrata Bobby
Barnes, Bradley
Barnes, Stuart
Adams
Adamsle
Ahmedper
Ahmed
Aliab
Ali
Aliyu
Alvarez
Alvarezte
Andersenmet
Andersen
Andersenpou
Andersonhel
Anderson
Andrewsjon
Andrews
Anthony
Anthonyjan
Antonella
Antonellacu
Armario
Armarioju
Arnoldar
Arnold
Backhauschri
Backhaus
Bakerell
Bakermi
Bakersu
Baker
Bal
Balre
Banerjee
Banerjeesub
Barnes
Barnesstu
79
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Bauer, András
Bauer, Hans H.
Bauer, Martina
Becker, Florian
Becker, Jan
Becker, Jochen
Becker, Katja
Becker, Kip
Bell, David
Bell, Pauline
Bell, Simon
Bengtsson, Anders
Bengtsson, Maria
Benito, Gabriel R.
G.
Benito, Leandro
Bennett, Dag
Bennett, Rebekah
Bennett, Roger
Black, Iain
Black, Nancy J.
Brady, John Patrick
Brady, Mairead
Brennan, Mike
Brennan, Ross
Broderick, Amanda
J.
Broderick, Anne
Broderick, Leigh J.
Brown, James R.
Brown, Jennifer
Brown, Joanne
Brown, Steven P.
Brown, UrsulaSigrid
Byrne, Angela
Byrne, Gabriel J.
Carneiro, Jorge
Carneiro, Maria João
Castro, Alberto Joao
C. De
Castro, Andres
Mazaira
Castro, Luciano
Thomé
Castro, Luís m. de
Chan, Jit
Chan, Priscilla Y. I.
Chan, Ricky Y. K.
Bauerand
Bauer
Bauermar
Becker
Beckerjan
Beckerjoc
Beckerkat
Beckerkip
Bellda
Bellpau
Bell
Bengtsson
Bengtssonmar
Benitogab
Benito
Bennettdag
Bennettre
Bennett
Blackia
Black
Brady
Bradymai
Brennan
Brennanro
Broderick
Broderickan
Brodericklei
Brown
Brownjen
Brownjoa
Brownste
Brownur
Byrne
Byrnegab
Carneiro
Carneiromar
Castroal
Castro
Castroluci
Castrolu
Chanjit
Chan
Chanri
80
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Chan, Tze Wee
Chandon, Jean-Louis
Chandon, Pierre
Chen, Bo-Wei
Chen, Chen-Yueh
Chen, Chien-Hung
Tom Chen
Chen, Fu-Tang
Chen, Kuang-Wei
Chen, Shu-Ching
Chen, Yuxin
Chen, Zhimin
Chien, Charles
Chien, Pi-Hsuan
Monica
Chou, Ting-Jui
Chou, Ting-Ting
Christensen, Alice
Slater
Christensen, Anne
Marie
Christensen, Bjørn
Christensen, Lars
Bech
Christensen, Sverre
Riiss
Claro, Danny
Pimentel
Claro, Priscila B.
Oliveira
Clark, Lindie
Clark, Moira
Clarke, Gary
Clarke, Ian
Coelho, Arnaldo
Coelho, Filipe
Cornelissen, Joep P.
Cornelissen, Markus
Costa, Claudia
Costa, Filipe
Campelo Xavier Da
Costa, Jorge
Cowley, Elizabeth
Cowley, Kym
Cox, David
Cox, Tony
Dahan, Ely
Dahan, Mariana
Medvetchi
Chantze
Chandonjea
Chandon
Chenbo
Chen
Chenchi
Chenfu
Chenkuan
Chenshu
Chenyu
Chenzhi
Chien
Chienpi
Chou
Chouti
Christensenal
Christensen
Christensenbjø
Christensenlar
Christensensver
Claro
Claropri
Clark
Clarkmo
Clarke
Clarkeian
Coelhoarn
Coelho
Cornelissen
Cornelissenmar
Costaclau
Costafi
Costa
Cowley
Cowleykym
Coxda
Cox
Dahan
Dahanmari
81
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Da Silva, Danielle
Mantovani Lucena
Da Silva, Helenita
R.
Da Silva, Jorge
Ferreira
Daskou, Antonia
Daskou, Sofia
Davies, Andrea
Davies, Iain A.
Dawes, John
Dawes, Philip L.
Denis, Jean-Emile
Denis, Sylvain
Derbaix, Christian
Derbaix, Maud
Devlin, Elinor
Devlin, James F.
De Wet, Dries
De Wet, Thinus
Dias, Campos
Roberta
Dias, José G.
Diehl, Kristin
Diehl, Sandra
Dimitriadis, Nikos
Dimitriadis, Sergios
Elliott, Richard
Elliott, Statia
Esteban, Águeda
Esteban, Mercedes
Evans, Jody
Evans, Martin
Falkenberg, Andreas
Falkenberg, Loren
Farrell, Colin
Farrel, Mark
Fernandes, Daniel
Fernandes, Joana
Cosme
Fernandez, Angel
Nogales
Fernández, Estela
Fernández, Pilar
Ferrín
Fernandez, Guillen
Fernandez,
Margarita
Ferreira, Alcina
Da Silvadan
Da Silva
Da Silvajor
Daskouan
Daskou
Davies
Daviesia
Dawesjo
Dawes
Denisjea
Denis
Derbaix
Derbaixma
Devlinel
Devlin
De Wet
De Wetthi
Dias
Diasjo
Diehlkri
Diehl
Dimitriadis
Dimitriadisser
Elliott
Elliottsta
Estebanágu
Esteban
Evans
Evansmar
Falkenberg
Falkenberglo
Farrellco
Farrell
Fernandes
Fernandesjo
Fernandezan
Fernándezest
Fernández
Fernandezgui
Fernandez
Ferreiraalc
82
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Teresa Gaspar
Ferreira, Armando
Leite
Ferreira, Jorge
Ferreira, Marco
Fischer, Bettina
Fischer, Marc
Fischer, Wolfgang
Fletcher, Keith
Fletcher, Richard
Fournier, Christophe
Fournier, Susan
Frank, Halsig
Frank, Lauri
Fry, Marie-Louise
Fry, Tim
Fu, Guoqun
Fu, Isabel
Fuchs, Christoph
Fuchs, Dion
Fuchs, Sebastian
Gabrielsen, Gorm
Gabrielsen, Tommy
Ganesan, Rama
Ganesan, Shankar
Gao, Hongzhi
Gao, Yuhui
Garcia, Christina
García, Iñaki
García, Irene
Garcia, José Luis
Garcia, Rosanna
García, Rosario
García, Teresa
Garnier, Aime
Isabelle
Garnier, Marion
Garrett, Jason
Garrett, Tony
Geiger, Ingmar
Geiger, Susi
Gomez, Jaime
Gomez, Jorge
Gomez, Miguel
Angel
Gomez, Monica
Suarez
Gomez, Pierrick
González, Celina
Ferreiraarm
Ferreira
Ferreiramar
Fischerbe
Fischer
Fischerwol
Fletcher
Fletcherri
Fournier
Fourniersu
Frank
Franklau
Frymar
Fry
Fu
Fuis
Fuchschri
Fuchs
Fuchsseb
Gabrielsengo
Gabrielsen
Ganesanra
Ganesan
Gaohong
Gao
Garcia
Garcíaiñ
Garcíair
Garciajos
Garciaro
Garcíarosa
Garcíater
Garnier
Garniermar
Garrettja
Garrett
Geigering
Geiger
Gomezjai
Gomezjor
Gomezmi
Gomezmon
Gomez
Gonzálezcel
83
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Gonzalez, José
González, Elisa Alén
Gonzalez, Maria
Victoria Roman
González, Óscar
González, Santiago
González, Varela
Gopalakrishna, Pillai
Kishore
Gopalakrishna,
Srinath
Graham, John
Graham, Stuart
Grant, David
Grant, Ian
Griffin, Abbie
Griffin, Deborah
Gruber, Barbara
Gruber, Thorsten
Haase, Kerstin
Haase, Nora
Hansen, Flemming
Hansen, Jonas Eder
Hansen, Kåre
Hansen, Lotte
Yssing
Hansen, Søren Sten
Hansen, Torben
Hansen, Felix
Harris, Fiona
Harris, Jennifer
Harris, Kim
Harris, Lloyd C.
Harris, Patricia
Harris, Phil
Hartmann, Adriane
Hartmann, Benjamin
Julien
Hartmann, Evi
Hartmann, Patrick
Heide, Morten
Heide, Jan
Helm, Roland
Helm, Sabrina
Hernández, Carlos
Hernández, Rosa
Herrmann, Andreas
Herrmann, Jean-Luc
Horváth, Csilla
Gonzalezjo
Gonzálezeli
Gonzalezmar
Gonzálezos
González
Gonzálezvar
Gopalakrishnapil
Gopalakrishna
Graham
Grahamstu
Grant
Grantian
Griffinab
Griffin
Gruberbar
Gruber
Haase
Haaseno
Hansenfle
Hansenjo
Hansen
Hansenlot
Hansensør
Hansentor
Hansenfel
Harrisfi
Harrisjen
Harriskim
Harrislloy
Harris
Harrisphi
Hartmann
Hartmannben
Hartmannevi
Hartmannpat
Heidemo
Heide
Helmro
Helm
Hernández
Hernándezro
Herrmannan
Herrmann
Horváth
84
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Horváth, Dóra
Hoffmann, Arvid
Hoffmann, Jonas
Hoffmann, Stefan
Hogg, Gill
Hogg, Margaret K.
Hoppe, Daniel
Hoppe, Melanie
Huber, Frank
Huber, JanAlexander
Hubert, Marco
Hubert, Mirja
Hughes, Mathew
Hughes, Paul
Hunter, Erik
Hunter, Louise E.
Iyer, Easwar S.
Iyer, Gopalkrishnan
Jaakkola, Elina
Jaakkola, Matti
Jiang, Minghua
Jiang, Zixi
Jimenez, Ana Isabel
Zarco
Jiménez, David
Jiménez, Julio
Jiménez, Nadia
Johanson, Jan
Johanson, Martin
Johansson, Johny
Johansson, Ulf
Johnson, Camille
Johnson, Eric J.
Johnson, Jean L.
Johnson, Michael D.
Jones, Jones
Jones, Marilyn Y.
Jones, Richard
Jones, Rosalind
Jones, Sandra C.
Jung, Kathrin
Jung, Susan
Kahn, Kenneth B.
Kahn, Mubbsher
Munawar
Kaiser, Christian
Kaiser, Jonas
Kaiser, Ulrike
Horváthdo
Hoffmannar
Hoffmannjon
Hoffmann
Hogggi
Hogg
Hoppedan
Hoppe
Huber
Huberjan
Hubert
Hubertmi
Hughes
Hughespa
Hunter
Hunterlou
Iyereas
Iyer
Jaakkolael
Jaakkola
Jiangmi
Jiang
Jimenezana
Jiménezda
Jiménezju
Jiménez
Johanson
Johansonmar
Johansson
Johanssonulf
Johnsonca
Johnsoner
Johnson
Johnsonmi
Jones
Jonesmar
Jonesri
Jonesro
Jonessan
Jungkath
Jung
Kahnken
Kahn
Kaiser
Kaiserjon
Kaiserul
85
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Kang, Seung-Mo
Kang, Suk-Hou
Kaya, Berhan
Kaya, Maria
Kefi, Samy
Kefi, Zied
Keller, Thomas
Keller, Veronika
Kemp, Gill
Kemp, Ron
Kent, Anthony
Kent, Raymond A.
Kim, Dong Ryul
Kim, Eun Young
Kim, Hye-ran
Kim, Hyoung Gil
Kim, Jaehwan
Kim, Jonathan
Kim, Jong Ho
Kim, Jungkeun
Kim, Jung Kyun
Kim, Ju-Young
Kim, Keysuk
Kim, Kwang-Seok
Kim, Kyung Hoon
Kim, Myung Soo
Kim, Namwoon
Kim, Pan Joon
Kim, Stephen
Kim, Yevgeniay
Kim, Young-man
Klaus, Martin
Klaus, Phil
Klein, Alex
Klein, Kristina
Klein, Maren
Knight, Dee K.
Knight, John
Koch, Christof
Koch, Jochen
Kozak, Metin
Kozak, Robert A.
Kumar, Alok
Kumar, Nirmalya
Kumar, Rajesh
Kwon, Jun Hyuk
Kwon, Ohjin
Lages, Carmen
Kangseu
Kang
Kayaber
Kaya
Kefisa
Kefi
Kellertho
Keller
Kempgi
Kemp
Kentan
Kent
Kimry
Kimeun
Kimhye
Kimgil
Kimjae
Kimjo
Kimho
Kimjung
Kimkyun
Kim
Kimkey
Kimkwa
Kimhoon
Kimsoo
Kimnam
Kimpan
Kimsteph
Kimyev
Kimyou
Klausmar
Klaus
Klein
Kleinkri
Kleinma
Knightdee
Knight
Kochchri
Koch
Kozak
Kozakro
Kumar
Kumarnir
Kumarra
Kwon
Kwonoh
Lages
86
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Lages, Cristina
Raquel
Lages, Luis
Lages, Natalia
Laroche, Mariane
Laroche, Michel
Laroche, Patrice
Larsen, Gretchen
Larsen, Ketil
Faldmark
Larsen, Nils Magne
Laurent, Florés
Laurent, George
Laurent, Gilles
Lee, Angela Y.
Lee, Bernard
Lee, Chang Han
Lee, Christina
Lee, Don
Lee, Dong-Hae
Lee, Hsin-Hsuan
Lee, Hye Jung
Lee, Hyun Seung
Lee, Janghyuk
Lee, Jimmy
Lee, Jong-Ho
Lee, Jong-Hwan
Lee, Jooyun
Lee, Ka
Lee, Michael S. W.
Lee, Nicholas
Lee, Nick
Lee, Seung-Hee
Lee, Seung-Hwan
Lee, Wonkyong
Lee, Zoe S.
Leonidou,
Constantinos
Leonidou, Leonidas
Lewis, Barbara R.
Lewis, Christopher
Lewis, Tony
Lim, Elison
Lim, Lynn L. K.
Liu, Ben ShawChing
Liu, Hong
Liu, Jia
Liu, Martin Jen-
Lagesra
Lageslu
Lagesna
Larochemar
Larochemi
Laroche
Larsengre
Larsenke
Larsen
Laurentflo
Laurentgeo
Laurent
Leeang
Leeber
Leecha
Leechri
Leedon
Leedong
Leehsin
Leehye
Leehy
Leeja
Lee
Leejoho
Leejohw
Leejoo
Leeka
Leemic
Leenich
Leeni
Leehee
Leeseu
Leewon
Leezo
Leonidoucon
Leonidou
Lewisba
Lewis
Lewiston
Limel
Lim
Liuben
Liuho
Liujia
Liumar
87
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Yuan
Liu, Pei Fen
Liu, Raymond
Liu, Rebecca
Liu, Tingchi
Lopez, Carmen
Lopez, Jose Angel
Sanchez
López, Inés
López, Moreno
Lorenza
Lopéz, Nicolas
Carolina
López, Pilar
Lopez, Raquel
Antolín
Lu, Irene R. R.
Lu, Junxiang
Lu, Vinh
Luna, David
Luna, Paula
Lutz, Antje
Lutz, Salla
Ma, Katherine H. Y.
Ma, Shan-Lyn
Ma, Yu
Ma, Zhenfeng
Marques, Alzira
Marques, Catarina
Marques, Susana
Martin, Brett
Martin, Christopher
Martin, Dan
Martin, Oscar
Martin, Pedro Juan
Martin, Xavier
Martínez, Carlos
Antonio
Martinez, Carole
Martínez, Eva
Martínez, Francisco
José
Mattsson, Jan
Mattsson, LarsGunnar
McDonald, Heath
McDonald, Seonaidh
Meier, Doreen
Meier, Helena
Liupei
Liuray
Liureb
Liu
Lopezcar
Lopez
Lópezin
López
Lopéznic
Lópezpil
Lopezra
Luir
Lu
Luvi
Luna
Lunapau
Lutzan
Lutz
Makath
Mashan
Ma
Mazhen
Marquesal
Marquescat
Marques
Martin
Martinchri
Martindan
Martinosc
Martinped
Martinxav
Martínez
Martinezcaro
Martínezeva
Martínezfran
Mattsson
Mattssonlar
McDonald
McDonaldseo
Meierdo
Meier
88
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Melnyk, Valentina
Melnyk, Vladimir
Menezes, João
Menezes, Rui
Michaelis, Lea
Michaelis, Manuel
Michel, Géraldine
Michel, Stefan
Miller, Ken
Miller, Klaus
Miller, Rohan
Mizerski, Dick
Mizerski, Kate
Moeller, Lise
Moeller, Timo
Moeller, Sabine
Molina, Arturo
Molina, Castillo
Francisco José
Morgan, Neil
Morgan, Robert
Morton, Fiona Scott
Morton, Peta
Mousley, Ben
Mousley, Wendy
Möller, Jana
Möller, Kristian
Mueller, Barbara
Mueller, Rene D.
Mueller, Steffen
Müller, Brigitte
Müller, Melanie
Nagy, Gabor
Nagy, Szabolcs
Nancarrow, Clive
Nancarrow, Pamela
Navarro, Angeles
Navarro, Antonio
Nelson, Charlie J.
Nelson, James
Nelson, Leif
Nelson, Michelle R.
Neumann, Debra
Neumann, Marcus
M.
Nguyen, Cathy
Nguyen, Tho
Nguyen, Trang
Melnyk
Melnykvla
Menezes
Menezesrui
Michaelis
Michaelisman
Michel
Michelste
Millerken
Miller
Millerro
Mizerskidi
Mizerski
Moellerlis
Moellerti
Moeller
Molinaart
Molina
Morganne
Morgan
Mortonfi
Morton
Mousleyben
Mousley
Möllerja
Möller
Muellerbar
Muellerre
Mueller
Müller
Müllermel
Nagy
Nagysza
Nancarrow
Nancarrowpam
Navarroang
Navarro
Nelson
Nelsonja
Nelsonle
Nelsonmich
Neumann
Neumannmar
Nguyencat
Nguyen
Nguyentra
89
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Nielsen, Brian
Nielsen, Anne
Ellerup
Olsen, G. Douglas
Olsen, Lars Erling
Palmer, Adrian
Palmer, Roger
Patterson, Maurice
Patterson, Paul
Paul, Gordon W.
Paul, Michael
Pauwels, Koen
Pauwels, Pieter
Peattie, Ken
Peattie, Sue
Pereira, Hélia
Goncalves
Pereira, Rosaria
Pereira, Teresa
Peters, Kim A.
Peters, Philipp
Polo, Peña Ana
Isabel
Polo, Yolanda
Poulis, Efthimios
Poulis, Konstantinos
Proenca, João F.
Proenca, Reinaldo
Rajh, Edo
Rajh, Suncana Piri
Ramanathan, Anand
Ramanathan, Ram
Rao, Sally
Rao, Vithala R.
Reed, Americus
Reed, Gary
Reid, Mike
Reid, Susan
Reinhold, Michael
Reinhold, Stephan
Reynolds, Kate
Reynolds, Kristy E.
Reynolds, Nina L.
Rivas, Eduardo
Rivas, Javier Alonso
Rivera, Jaime
Rivera, Pilar
Robben, John
Robben, Roderick
Nielsen
Nielsenan
Olsen
Olsenlar
Palmer
Palmerro
Pattersonmau
Patterson
Paulgor
Paul
Pauwels
Pauwelspie
Peattie
Peattiesue
Pereirahél
Pereiraro
Pereira
Peterski
Peters
Polo
Poloyo
Poulis
Pouliskon
Proenca
Proencarei
Rajh
Rajhsun
Ramanathan
Ramanathanram
Raosa
Rao
Reedam
Reed
Reidmi
Reid
Reinhold
Reinholdste
Reynolds
Reynoldskri
Reynoldsni
Rivas
Rivasjav
Riverajai
Rivera
Robben
Robbenrod
90
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Robson, Andrew
Robson, Matthew
Rodrigues, Luiza
Rodrigues, Paula
Rodríguez, Ana
Isabel
Rodríguez, Manuel
Rodríguez, Molina
Miguel Angel
Rodríguez, Pinto
Javier
Rodríguez, Rosa
María
Romero, Carlota
Lorenzo
Romero, Jaime
Rose, Gregory M.
Rose, John
Rossi, Carlos
Alberto Vargas
Rossi, Peter
Roth, Katharina
Petra
Roth, Stefan
Rowe, Amanda T.
Rowe, Susan
Rubio, Ana Garrido
Rubio, Luis
Rubio, Natalia
Ruiz, Carla
Ruiz, David Martin
Ruiz, Enar
Ruiz, Molina Maria
Eugenia
Ruiz, Rocio Rangel
Ruiz, Salvador
Russell, Cristel
Russell, Dale
Sackett, Anna
Sackett, Aron
Saker, Jim
Saker, Merav
Salavou, Eleni
Salavou, Helen
Salazar, Ana
Salazar, Maria T.
Sánchez, Isabel
Sánchez, Javier
Sánchez, Manuel
Robson
Robsonmat
Rodrigues
Rodriguespau
Rodríguezana
Rodríguezma
Rodríguez
Rodríguezpin
Rodríguezrosa
Romerocar
Romero
Rose
Rosejo
Rossicar
Rossi
Rothkat
Roth
Rowe
Rowesu
Rubioana
Rubiolu
Rubio
Ruiz
Ruizda
Ruizen
Ruizmol
Ruizro
Ruizsa
Russell
Russellda
Sackettan
Sackett
Saker
Sakerme
Salavou
Salavouhel
Salazar
Salazarmar
Sánchezisa
Sánchezja
Sánchezma
91
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Sánchez, Raquel
Sánchez, Teresa
Santos, Cristiane
Santos, Jessica
Santos, José
Santos, Libia
Santos, Maria Da
Conceicão
Santos, Mirela
Santos, Rubens Da
Costa
Santos, Vijande
Maria Leticia
Schmidt, Marcus
Schmidt, Svenja
Schmidt, Thomas
Schmidt, Thorsten
Schmidt, Volker
Schmitt, Julien
Schmitt, Philipp
Schmitt, Robert
Schulze, Caroline
Schulze, Timo
Scott, Adrianne
Scott, Jane
Sharma, Neha
Sharma, Neeru
Sharp, Anne
Sharp, Byron
Shaw, Deirdre
Shaw, Mike
Shaw, Robin N.
Shaw, Vivienne
Shin, Jong-Kuk
Shin, Meongijn
Silva, Carla
Silva, Catarina
Silva, Maria Jose
Simon, Henrik
Simon, Judith
Simon, Steven J.
Sinclair, Julie
Sinclair, Thea
Singh, Jagdip
Singh, Jaywant
Singh, Ramendra
Singh, Satyendra
Singh, Siddharth
Sinha, Ashish
Sánchez
Sánchezter
Santoscri
Santosjes
Santosjo
Santosli
Santosmar
Santosmi
Santosru
Santos
Schmidtmar
Schmidt
Schmidttho
Schmidtthor
Schmidtvol
Schmitt
Schmittphi
Schmittro
Schulzecar
Schulze
Scottad
Scott
Sharma
Sharmanee
Sharp
Sharpby
Shawde
Shawmi
Shawro
Shaw
Shinjo
Shin
Silva
Silvacat
Silvamar
Simon
Simonju
Simonste
Sinclair
Sinclairthe
Singhjag
Singhjay
Singh
Singhsaty
Singhsid
Sinhaas
92
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Sinha, Prabhakant
Slater, Rod
Slater, Stanley F.
Slater, Stephanie
Smit, Edith
Smit, Willem
Smith, Andrew
Smith, A. P.
Smith, Daniel C.
Smith, Gareth
Soares, Ana
Soares, Elsa
Sood, Ashish
Sood, Sanjay
Song, In Am
Song, Michael
Song, Sangyoung
Srivastava, Rajendra
Srivastava, Rajesh
Steiner, Michael
Steiner, Susanne
Steiner, Winfried J.
Sun, Daewon
Sun, Luping
Taylor, Charles R.
Taylor, Gail
Taylor, Paul
Teichmann, Karin
Teichmann, MaikHenrik
Thompson, Jennifer
Thompson, Yvonne
Tuominen, Matti
Tuominen, Sasu
Van Der Lans, Ivo
A.
Van Der Lans, Ralf
Van Dijk, Albert
Van Dijk, Gert
Villanueva, Julian
Villanueva, María
Luisa
Vogel, Johannes
Vogel, Verena
Voss, Glenn
Voss, Roediger
Voss, Zannie
Wagner, Janet
Wagner, Ralf
Sinha
Slaterrod
Slatersta
Slater
Smited
Smit
Smithan
Smithap
Smith
Smithga
Soares
Soaresel
Soodas
Sood
Songin
Song
Songsang
Srivastavarajen
Srivastava
Steinermi
Steinersus
Steiner
Sun
Sunlu
Taylorchar
Taylor
Taylorpa
Teichmannka
Teichmann
Thompson
Thompsonyv
Tuominen
Tuominensa
Van Der Lans
Van Der Lansra
Van Dijkal
Van Dijl
Villanueva
Villanuevamar
Vogeljo
Vogel
Vossgle
Voss
Vosszan
Wagnerjan
Wagner
93
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Wagner, Stephan M.
Wagner, Udo
Walker, Joan L.
Walker, Rhett
Walter, Achim
Walter, Eva
Wang, I Chen
Wang, Lai-Wang
Wang, Qing
Wang, Weiyue
Wang, Yantao
Ward, Janet
Ward, Ronald W.
Weber, Bernd
Weber, Karin
Wilke, Annika
Wilke, Ricky
Wilke, Sina
Wilson, Alan
Wilson, Brad
Wilson, Elaine
Wilson, Juliette
Wong, Ipkin
Wong, Mandy
Wong, Nancy
Wong, Veronica
Wright, George
Wright, Len Tiu
Wright, Malcom
Wright, Owen
Wu, Steven
Wu, Wann-Yih
Xie, Chunyan
Xie,Ying
Yang, Rae Jin
Yang, Yinghui
Young, J. A.
Young, Karen
Young, Louise
Young, Scott H.
Yu, Tiffany HuiKuang
Yu, Ting
Yu, Wantao
Zhang, Hongbo
Zhang, Jing
Zhang, John
Zhang, Lida
Wagnerste
Wagnerudo
Walkerjo
Walker
Walterach
Walter
Wangi
Wanglai
Wang
Wangwe
Wangyan
Wardja
Ward
Weberbe
Weber
Wilkean
Wilke
Wilkesi
Wilsonal
Wilsonbrad
Wilson
Wilsonju
Wong
Wongma
Wongnan
Wongve
Wrightgeo
Wrightlen
Wright
Wrightow
Wuste
Wu
Xiechu
Xie
Yang
Yangra
Youngja
Youngka
Younglou
Young
Yutif
Yu
Yuwan
Zhanghong
Zhangji
Zhang
Zhangli
94
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Zhang, Ting
Zhu, Judy Li
Zhu, Tian
Zhangti
Zhu
Zhuti
95
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Appendix 2: Track List
Tracks as listed in the EMAC conference proceedings.
Track 1:
Advertising, Promotion and Marketing Communication
Track 2:
Business-to-Business Marketing and Networks
Track 3:
Consumer Behaviour
Track 4:
Innovation and New Product Development
Track 5:
International and Cross-Cultural Marketing
Track 6:
Marketing in Emerging and Transition Economies
Track 7:
Marketing of Public and Non-Profit Organisations
Track 8:
Marketing Research and Research Methodology
Track 9:
Marketing Strategy and Leadership
Track 10:
Marketing Theory
96
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Track 11:
Modeling and Forecasting
Track 12:
New Technologies and E-Marketing
Track 13:
Pricing and Financial Issues in Marketing
Track 14:
Product and Brand Management
Track 15:
Relationship Marketing
Track 16:
Retailing, Channel Management and Logistics
Track 17:
Sales Management and Personal Selling
Track 18:
Service Marketing
Track 19:
Social Responsibility, Ethics and Consumer Protection
Track 20:
Tourism Marketing
97
A Social Network Analysis of the EMAC Annual Conferences 2000-2010
Appendix 3: Data CD
98