Managers and members in online communities of practice: What are they talking about?

Abstract

Online communities of practice are becoming important organizational arenas. Much literature about online communities describes them as environments based on user-generated content. This article shifts the focus from community members to the managers, demonstrates their central role in creating content and promoting discussions, and distinguishes between managers and members by asking: Who posts more? What themes do the posts of managers and members focus on? Who is more and who is less critical of the communities and practice? Who shares more of their personal experience? And finally, Who asks questions and who provides answers?

Full text

Managers and Members in Online Communities of
Practice: What are they Talking About?
Azi Lev-On
School of Communication
Ariel University
Ariel, Israel
+972(0)39371439
azilevon@gmail.com
Nili Steinfeld
School of Communication
Ariel University
Ariel, Israel
+972(0)39371439
nilisteinfeld@gmail.com
ABSTRACT
Online communities of practice are becoming important
organizational arenas. Much literature about online communities
describes them as environments based on user-generated content.
This article shifts the focus from community members to the
managers, demonstrates their central role in creating content and
promoting discussions, and distinguishes between managers and
members by asking: Who posts more? What themes do the posts
of managers and members focus on? Who is more and who is less
critical of the communities and practice? Who shares more of
their personal experience? And finally, Who asks questions and
who provides answers?
General Terms
Management, Measurement, Human Factors
Keywords
Communities, Communities of Practice, Online Communities of
Practice, Management, Content Analysis, Digital Government, e-
Government, Social Media, Social Work
1. INTRODUCTION
Governance systems are characterized by 'horizontal' and 'vertical'
dimensions of command and control (Sartori, 1987). Online as
well as offline communities attract public and scholarly attention
due to the emphasis on their ‘horizontal’ dimension of
governance, which is expressed by peer production, monitoring,
and sanctioning of members of the community in a collaborative
manner (Connolly, 2001; Kim, 2000; Kollock, 1999; Komito,
1998; Preece, 2000; Resnick, 2002; Wellman & Gulia, 1999). But
there is also a ‘vertical’ dimension to communities, which may be
prominent in online even more than in offline communities.
Online communities have owners, technical managers, designers,
and moderators who allow the existence of the community and
perform activities essential to the creation and preservation of the
platform and framework of the discussions around which the
community evolves (Butler et al., 2007; Kim, 2000; Preece,
2000).
However, despite the importance of the vertical dimension of
online communities, the majority of research focuses on their
horizontal dimension (Fallah, 2011). This study helps fill that void
by comparing the scope and character of content created by
community managers with the content created by ordinary
members.
The use of online platforms to create and maintain a community
poses considerable challenges to community managers. Unlike
leaders of groups that meet face to face, online community
managers are not physically present, cannot solve problems face
to face, and cues that may emphasize their authority are dimmed
(Kayworth & Leidner, 2002). Moreover, every discussion that the
managers are involved in is stored online, which makes the
analysis of their content much easier, but limits the flexibility of
managers as their online statements and decisions are all open and
transparent.
Still, the literature demonstrates that community managers and the
functions that they typically perform have a major impact on the
discourse in the community, on the realization of community
goals and its success in terms of recruitment and member
management, content creation, and more (Bishop et al., 2008;
Bourhis, Dubé & Jacob, 2005). The little available literature
addresses the typical roles of community managers, among them:
• Member management – recruiting members through
advertising, public relations, or other ways; removing
members if necessary; encouraging members to perform
certain activities for the community (Bateman, 2008; Butler
et al., 2007; Osimo, 2008).
• Content management – supervising the agenda of the
community, encouraging discussions while avoiding
information overload, preventing "flaming", and removing
inappropriate posts (Meier, 1997; Osimo, 2008). Managers
may also initiate special events, such as discussions with
experts (Kim, 2000).
• Establishing and maintaining rules and norms – Clarifying
group norms to new and existing members, including
expected conducts and sanctions, resolving disputes, and if
necessary removal of members whose behavior does not
conform to the rules (Kim, 2000; Preece, 2000).
For community members too, the manager's function is perceived
as critical to the success of the community, and her actions
contribute to the transformation of a platform for exchange of
information, to a community of practice – a space where
knowledge is built through shared learning (Gray, 2004).
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. Copyrights
for components of this work owned by others than ACM must be
honored. Abstracting with credit is permitted. To copy otherwise, or
republish, to post on servers or to redistribute to lists, requires prior
specific permission and/or a fee. Request permissions from
Permissions@acm.org.
dg.o '14, June 18 - 21 2014, Aguascalientes, Mexico
Copyright 2014 ACM 978-1-4503-2901-9/14/06…$15.00.
http://dx.doi.org/10.1145/2612733.2612771

The limited academic literature about the functioning and
influence of managers of online communities is based on
interviews or small group experiments. Although existing studies
have revealed that a manager's role is critical in the community,
they do not demonstrate quantitatively how these roles are
reflected in the dynamics of the community, and they do not focus
on the content of posts published by managers. This is the first
empirical study, to our knowledge, that focuses on the content
posted by community managers and does so through content
analysis from online communities of practice, and on a large
quantity of data (7,248 posts in 11 communities of practice).
Thus, this article lays the ground for relevant and innovative
investigation of the functioning of managers of online
communities of practice.
2. METHODOLOGY
The research arena of the current study is the communities of
practice established by the Israel Ministry of Social Affairs and
Social Services. The project was launched in late 2006 with the
goals of developing methodologies of organizational learning
among social workers, promoting learning in different
organizations, and utilizing ICTs in the social services (Fein,
2011; Sabah, 2010). At the time of data collection for the study
(early 2012), 31 communities were active on the ministry's site
with more than 7700 registered members. Although the
communities are hosted on the servers of the Ministry of Social
Affairs, less than 30% of the members are employees of the
ministry, and the communities also involve employees of
municipalities, NGOs, legal councils, and other stakeholders.
Thus, the communities bring together professionals to address
issues related to the social services, and function as an arena for
encounters between different stakeholders involved in similar
endeavors.
Members of the communities are identified according to their
name and professional role. The member list is public and
available to all members of the community, so everyone knows
who might read and comment on content uploaded to the
community. The community managers have volunteered for their
role, and are members of the communities themselves. They
receive token payment in the form of vouchers (Sabah, 2010).
The present study focuses on 11 out of 31 communities of practice
at the time of data collection. The communities that were selected
cover a range of variation of relevant variables for the research, to
increase internal and external validity of the study:
• The date of the establishment of the community (new
communities versus older communities).
• Level of activity (high, medium, low).
• The size of the community (large, medium, small).
• Different professional domains (therapeutic communities
compared to communities dedicated to procedures and
administrative issues).
The sample included the following 11 communities: Intellectual
Disability, Community Work, Children at Risk, Immigrants and
Inter-Cultural Issues, Policy and Performance, Blind and the
Visually Impaired, Domestic Violence, Welfare Management in
Municipalities, Foster Care, Organizational Learning, and
Juvenile Delinquency. In each community of practice all posts
available at the time of data collection were analyzed. A total of
7248 posts were coded using a coding sheet developed for the
study.
The study involves two units of analysis: single posts and
threaded discussions (a first post and at least one additional
comment related to it). Thus, some of the categories in the coding
sheet relate to posts and others to discussions.
Altogether, the coding sheet comprised 24 quantitative categories,
which included the identity of the author (manager or community
member), time of publication, the name of the community where
the post was published, the post's order (is it a first post, opening a
new discussion or a response to a previous post), and a number of
categories relevant to the content of the post, as will be described
later in the paper. The posts were coded by 13 coders after an
intensive training and a reliability test, which was repeated until a
reliability of 90% between the coder and the leading researcher
was obtained in each of the categories of the coding book. In
addition, the first set of posts each coder has coded (usually 50-
120 posts) were examined by the lead researcher, who reviewed
the coding to make sure that reliability is maintained.
In the current study, 308 first posts (opening a new discussion)
were published by managers of the communities, and 1,201 first
posts were published by members. In addition to the analysis of
all posts, first posts were analyzed separately. Underlying this
decision was the assumption that if community managers behave
differently than the rest of the members it may be reflected
especially in messages that open a discussion, where a new topic
is introduced and a discussion is initiated, rather than when posts
are published in response to previous posts.
3. HYPOTHESES
At the near absence of relevant academic literature focusing on
the content posted by managers of online communities of practice,
the hypotheses of this study are exploratory in nature. The
hypotheses are:
Content Posted by Managers and Members
H1: First posts (opening new discussions) by managers address
different themes than posts by members. Due to their role and
place in the organization (mainly in the online community itself),
managers address organizational issues more than members, while
members focus on practical issues, related to their daily work.
H2: Similarly, in a more general analysis of all posts, posts by
managers address more organizational/ministerial issues, while
posts by members address more practical issues.
H3: Discussions following first posts by managers will also
address more organizational issues, since managers will tend to
develop discussions dealing with these issues. Member-initiated
discussions will address more practical issues.
H4: Sharing personal experiences: as members are more practice-
oriented in their posts, they share more personal experiences than
managers do.
H5: References to the communities of practice project: Since
managers are more organizational-oriented, they make more
references to the communities of practice project than ordinary
members do.
H6: Criticism: Since managers are more organization-oriented,
and may perceive themselves as representatives of the project and
of the ministry, they will be less judgmental and will express less
criticism than ordinary community members.

Questions and Answers
H7: Given their managerial role in the community, overseeing the
community's activity in general, community managers will be less
likely to include any form of assistance (by either requesting or
providing assistance) compared to ordinary community members.
H8: Due to a manager’s role in directing discussions and
promoting dialogue and exchange of knowledge between
members, when questions are raised in discussions that involve
managers, more relevant answers will be given than in
discussions among members only.
4. FINDINGS
4.1 Overview
The literature review demonstrates the importance attributed to
the role of the manager in online communities of practice, and the
perceived impact of her behavior over the dynamics in the
community. Some general statistics demonstrate this may apply to
the communities studied here as well. Community managers
posted 17.9% of the messages in the corpus, while 51.3% of all
posts belong to discussions that involved community managers as
well as ordinary members. We found that 23.8% of the posts by
managers (and 20.2% of the posts by members) are first posts,
26.6% of the posts by managers and 35.7% of the posts by
members are responses to first posts, while 49.6% of the posts by
managers and 44.2% of the posts by members are higher-order
responses (i.e., responses to responses to first posts, responses to
responses to responses to first posts, etc.).
4.2 Content Posted by Managers and
Members
Are there differences between the content contributed by
managers and members? Table 1 presents the contents of first
posts, posts in general, and discussions initiated by members and
managers. The categories that were included in the coding sheet
were: Practical advice – which is directly related to daily work
with clients, for example: what is the impact of certain kinds of
interventions?; Organizational advice – which is related to
employees’ daily work but unrelated to working with clients, for
example concerning forms, procedures, programs, and courses;
Academic advice – references to academic literature, relevant
research etc.; Emotional support – addressing community
members’ manifestations of charged emotions (anger, frustration,
fear, sadness, etc.) that are related to their work; Informing about
an event or conference; Greetings and gratitude; Publication of a
project or organization; Submitting contact details; Statements
about the community’s topic – related not to employees’ daily
work, but to more general issues related to the community’s main
theme, for example: How to improve service for patients, How to
improve the status of blind people in the Israeli society and finally
– Other topics.
Table 1. Topics of first posts, posts in general, and posts in discussions initiated by managers and members. Significance was
calculated using Chi-Square tests.
First Posts
All Posts
Discussions Initiated
Topic of post
% by manager
% by members
% by manager
% by members
% by manager
% by members
Practical advice
20.1%***
39%***
28.3%***
41.7%***
32.7%***
43.6%***
Organizational advice
26.6%
25.9%
29.1%***
25%***
27.9%*
25.2%*
Academic advice
9.7%
10.6%
9%***
5.8%***
4.5%***
6.6%***
Emotional support
4.5%
3.2%
3.8%
3.3%
3.5%
3.6%
Informing on an event or
conference
17.5%***
11.1%***
10.5%***
7.6%***
9.9%***
6.4%***
Greetings and gratitude
3.9%
3.7%
8.9%***
11.4%***
11%
11.6%
Publication of a project or
organization
23.4%***
15.1%***
12.1%***
7.7%***
8.5%*
6.9%*
Submitting contact details
6.2%*
10.1%*
2.9%***
8.6%***
1.5%***
8.8%***
Statements about the
community’s topic
21.1%***
10.3%***
21.5%
23.3%
27.4%***
22.9%***
Other topics
10.4%*
6.2%*
14.8%***
7.3%***
13.9%***
6.9%***
* p < .05, ***p < .01

The findings for first posts, posts in general, and discussions
demonstrate that the leading conversation categories for managers
and members alike are practical advice, organizational advice, and
statements about the community’s topic.
The findings support hypotheses H1-H3:
• First posts: First posts by members include a significantly
higher rate of practical advice, while posts by managers have
a higher percentage of statements about the community’s
topic and publication of events and projects. No significant
differences were found in terms of organizational advice,
academic advice, or emotional support.
• All Posts: The three most popular categories (practical
advice, organizational advice, and statements about the
community’s topic) are identical for managers and members.
This finding makes sense, given that managers and members
respond to each other and therefore engage in the same
issues. However, differences exist: members prefer to discuss
practical issues related to their day-to-day work significantly
more than managers (41.7% vs. 28.3%), while managers
prefer providing information of various kinds, and especially
organizational help – information related to formal training,
job openings, courses, forms needed to be submitted, etc.
Interestingly, emotional support is very rare among both
managers' and members' posts (only 3.8% of managers' posts
and 3.3% among members' posts).
• Discussions: 1897 posts were included in discussions that
followed from first posts by managers, while 5340 posts
were included in discussions that followed from first posts by
members. Messages in discussions initiated by members
include a higher rate of practical and academic advice, while
messages by managers have a higher rate of organizational
advice, statements about the community’s topic, and
publications of events or projects. No significant differences
between managers and members were found in terms of
emotional support.
4.3 Sharing Personal Cases
Members of the communities are more likely than managers to
share personal cases they encountered during their work (7.5% vs.
2.2%). A Chi-square test found a significant correlation between
the identity of the author (manager or member) and sharing
personal cases (χ2= 49.86, p<0.01). The effect size was calculated
using Cramer's V and was found to be weak (r=0.08). This finding
supports hypothesis H4.
4.4 Referring to the Communities of Practice
Project
Community managers are much more likely than other members
of the community to refer to the communities of practice project
in their posts. Reference to the communities of practice project
was seen in 26.1% of the posts by managers compared to 10.8%
of the posts by other members of the community.
However, the relative percentage of posts that include positive and
negative references towards the project is almost the same
between managers and members. In both groups the majority of
the references are positive, i.e., there is very little criticism, but
rather praise or simply a description of the uses (e.g., uploading
articles, forms, etc.).
A Chi-square test was performed to examine the relation between
the identity of the author of a message (managers vs. members)
and referencing the communities of practice project. The relation
between these variables was significant (χ2= 217.67, p<0.01). The
effect size was calculated using Cramer's V and was found to be
weak (r=0.17). Table 2 summarizes the distribution of references
to the project between managers and members of the
communities. This finding supports hypothesis H5.
Table 2. Reference to the communities of practice project by
managers and members of the communities.
Author
% of posts
that include
references
to
communities
of practice
% of
posts
including
positive
attitude
towards
the
project
% of
posts
including
mixed
attitude
towards
the
project
% of
references
including
negative
attitude
towards
the
project
Managers
26.1%
85.2%
11.83%
2.95%
Members
10.8%
83.59%
14.37%
2.03%
4.5 Criticism by Managers and Members
No statistically significant differences were found between the
percentage of community managers' posts containing criticism
and the percentage of posts by other members of the community
containing criticism. Community managers tend to be slightly less
critical than other members of the community (5.7% vs. 7.6%).
When criticism does appear it seems that community managers
tend to express less criticism towards state authorities (22.97% vs.
28%) and the Ministry of Social Affairs and Social Services
(16.21% vs. 21.19%) than other members of the communities.
Conversely, managers express more criticism towards members of
the community themselves than the rest of the community
members (24.32% vs. 14.56%). This finding does not lend support
to hypothesis H6.
4.6 Requesting and Providing Assistance
Do community managers' posts contain more or fewer requests
and provision of assistance than messages posted by community
members?
Let us note that nearly 40% of community managers' posts include
no references to providing assistance or requesting assistance,
compared to only 31.8% of the messages by members of the
communities. In order to test the relation between the identity of
the post’s author (manager or member) and assistance
request/provision in the respective posts, a chi-square test was
conducted. The relation between these variables was significant
(χ2= 57.94, p<0.01). The effect size was calculated using Cramer's
V and was found to be weak (r=0.09). Thus, managers are less
likely to include a form of assistance in their posts than other
members of the community, as hypothesis H7 suggests. Table 3
summarizes the results. As shown in the table, while managers
tend slightly more than members to request assistance in their
posts (either for them or for other community members), they are
much less likely to provide assistance in their posts.

Table 3. Percentage of messages including provision or
requests for assistance, by author of the post.
Author
% of
messages
including
requests for
assistance
% of
messages
including
provision of
assistance
% of
messages
including
requests and
provision of
assistance
% of
messages
that include
neither
requests nor
provision of
assistance
Manager
24.6%
32.8%
2.8%
39.8%
Member
21.5%
44.2%
2.5%
31.8%
4.7 Questions and Answers
Lastly, do discussions that managers participate in include
relatively more answers to questions raised within the discussion,
compared to discussions where only community members
participate?
To answer this question, from a smaller sample including only
discussions in which a question was raised, a chi-square test was
performed to examine the relation between the identity of the
participants (only managers, only members, managers and
members) and receiving an answer to the question raised in the
discussion. The relation between these variables was significant
(χ2= 176.11, p<0.01), supporting hypothesis H8. The effect size
was calculated using Cramer's V and was found to be moderate
(r=0.34). While in discussions including both managers and
members 85% of the questions were answered, only 74% of
questions raised in discussions including only members were
answered. Interestingly, in discussions that include managers
alone, less than a quarter of the questions raised within the
discussion are answered. Table 4 summarizes the results.
Table 4. Answering questions within discussions, by
participant identity.
Participants
in the
discussion
An answer to
the question
was not
provided
Unclear
whether a
relevant
answer was
provided
A relevant
answer was
provided
Only managers
68%
8%
24%
Managers and
members
3.1%
11.6%
85.3%
Only members
5.1%
20.9%
74%
5. DISCUSSION AND CONCLUSIONS
The findings call for a new conceptualization of communities of
practice. Instead of the common perception of CoPs as
environments sustained by user-generated content, the findings
clearly call for a new understanding of communities of practice as
environments orchestrated by the continuing dialectic between
managers and members.
The findings demonstrate that there are significant differences in
the behaviors of managers and members of communities of
practice in terms of content creation, topics discussed, and
involvement in asking questions and addressing others’ concerns.
All our exploratory hypotheses (with the exception of H6) were
supported. The results indicate a variety of functions of managers
of communities of practice.
5.1 The Manger as a Content Manufacturer
Community managers are very active in initiating discussions and
creating content. Managers wrote 17.9% of the posts published
within the 11 communities we analyzed; less than 40% of the
messages belong to discussions among only members of the
community.
5.2 The Manager as a Criticizing Authority
and less as an "Open Ear"
This study shows that despite a lot of involvement and
participation of the community managers, when it comes to
sharing personal cases, community members often share more
personal cases than managers. This may indicate that despite the
community manager being a significant factor in the community's
activity and content, he or she is still considered as part of the
executive level as far as personal issues are concerned. Another
finding that strengthens this argument is the existence of relatively
few messages that include an element of providing assistance or
requesting assistance. In other words, members of the
communities who are not managers tend to post more messages
containing provision of assistance. However, when it comes to
answering questions within a discussion, it seems that the
presence of a manager in a discussion tends to encourage
answering of questions raised within the discussion, may it be the
manager or a member of the community answering the question.
5.3 The Manger as an Organizational Mentor
The three most common topics of conversation among both
community managers and other members of communities are the
same: practical assistance, organizational assistance, and
expressing personal opinions on an issue. Still, findings suggest
that managers of communities of practice prefer providing
information of various kinds, but especially information related to
formal training, job vacancies, courses, forms, and other general
organizational and technical information that can help other
members. The organizational assistance occurs much more often
than practical assistance, which is essential in the context of day-
to-day work with clients. Community managers tend, much more
than other members, to refer to the communities of practice
project.
Future work can continue to explore the functioning of managers
of communities of practice as discussion moderators, study the
effects of interventions on their behalf, explore the manager role
in answering or raising questions that promote discussions, or in
handling conflicts and crises, for example, when a form of
criticism or complaint towards the organization or the community
is raised by members.
6. ACKNOWLEDGMENTS
The study was supported by the Center for the Study of New
Media, Society and Politics at Ariel University. The authors thank
Keren Sereno and Odelia Adler for their assistance in analyzing
the data and finalizing the manuscript.
7. REFERENCES
[1] Bateman, P. J. 2008. Online Community Referrals and
Commitment: How Two Aspects of Community Life Impact

Member Participation. Doctoral Dissertation, University of
Pittsburgh.
[2] Bishop, J., Bouchlaghem, D., Glass, J. and Matsumoto, I.
2008. Identifying and implementing management best
practice for communities of practice. Architectural
Engineering and Design Management, 4(3-4), 160-175.
[3] Bourhis, A., Dubé, L., and Jacob, R. 2005. The success of
virtual communities of practice: The leadership factor. The
Electronic Journal of Knowledge Management, 3, 1
(Jul.2005), 23-34.
[4] Butler, B., Sproull, L., Kiesler, S. and Kraut, R. 2007.
Community effort in online groups: Who does the work and
why? In Leadership at a Distance, S. Weisband Ed.
Lawrence Erlbaum Associates, New York, 171-194.
[5] Connolly, R. 2001. The rise and persistence of the
technological community ideal. In Online Communities:
Commerce, Community Action and the Virtual University, C.
Werryand M. Mowbray Eds. Upper Saddle River, Prentice
Hall, NJ, 317-364.
[6] Fallah, N. 2011. Distributed form of leadership in
communities of practice. International Journal of Emerging
Science, 1, 3 (Sep. 2011), 357-370.
[7] Fein, T. 2011. Online Communities of Practice in the Social
Services. MA Thesis, the Hebrew University. (Hebrew)
[8] Gray, B. 2004. Informal learning in an online community of
practice. Journal of Distance Education, 19, 1, 20-35.
[9] Kayworth, T. R., and Leidner, D. E. 2002. Leadership
effectiveness in global virtual teams. Journal of Management
Information Systems, 18, 3, 7-40.
[10] Kim, A.J. 2000. Community Building on the Web: Secret
Strategies for Successful Online Communities. Berkeley:
Peachpit Press.
[11] Kollock, P. 1999. The economies of online cooperation:
Gifts and public goods in cyberspace. In Communities in
cyberspace, M.A. Smith and P. Kollock Eds. Routledge,
New York, 220-239.
[12] Komito, L. 1998. The net as a foraging society: Flexible
communities. The Information Society, 14 (Apr-Jun.1998),
97-106.
[13] Meier, A. 1997. Inventing new models of social support
groups: A feasibility study of an online stress management
support group for social workers. Social Work with Groups,
20, 4, 35-53.
[14] Osimo, D. 2008. Web 2.0 in Government: Why and How?
Technical Report. Institute for Prospective Technological
Studies (IPTS), JRC, European Commission.
http://ftp.jrc.es/EURdoc/JRC45269.pdf
[15] Preece, J. 2000. Online Communities: Designing Usability
and Supporting Sociability. New York: Wiley.
[16] Resnick, P. 2002. Beyond bowling together: Socio Technical
capital. In HCI in the New Millennium, J.M. Carroll Ed.
Addison-Wesley, New York, 647-672.
[17] Sabah, Y. 2010. Online learning groups and communities of
practice in the social services. Society and Welfare, 30, 1,
111-130. (Hebrew).
[18] Sartori, G. 1987. The Theory of Democracy Revisited, v. 1.
Chatham House Publishers, Chatham, N.J.
[19] Wellman, B., and Gulia M. 1999. Virtual communities as
communities. In Communities in cyberspace, M.A. Smith
and P. KollockEds. Routledge, New York, 167-194.