Conference PaperPDF Available

Fixing the ‘Out of sight out of mind’ problem one year of mood-based microblogging in a distributed software team


Abstract and Figures

Distributed teams face the challenge of staying connected. How do team members stay connected when they no longer see each other on a daily basis? What should be done when there is no coffee corner to share your latest exploits? In this paper we evaluate a microblogging system which makes this possible in a distributed setting. The system, WeHomer, enables the sharing of information and corresponding emotions in a fully distributed organization. We analyzed the content of over a year of usage data by 19 team members in a structured fashion, performed 5 semi-structured interviews and report our findings in this paper. We draw conclusions about the topics shared, the impact on software teams and the impact of distribution and team composition. Main findings include an increase in team-connectedness and easier access to information that is traditionally harder to consistently acquire.
Content may be subject to copyright.
Fixing the ’Out of Sight Out of Mind’ Problem
One Year of Mood-Based Microblogging in a Distributed Software Team
Kevin Dullemond, Ben van Gameren
Delft University of Technology
The Netherlands
{k.dullemond, b.j.a.vangameren}
Margaret-Anne Storey
Univerity of Victoria
Arie van Deursen
Delft University of Technology
The Netherlands
Abstract—Distributed teams face the challenge of staying
connected. How do team members stay connected when they
no longer see each other on a daily basis? What should be done
when there is no coffee corner to share your latest exploits? In
this paper we evaluate a microblogging system which makes this
possible in a distributed setting. The system, WeHomer, enables
the sharing of information and corresponding emotions in a fully
distributed organization. We analyzed the content of over a year
of usage data by 19 team members in a structured fashion,
performed 5 semi-structured interviews and report our findings
in this paper. We draw conclusions about the topics shared,
the impact on software teams and the impact of distribution
and team composition. Main findings include an increase in
team-connectedness and easier access to information that is
traditionally harder to consistently acquire.
The year is 2011 and in a young and small software com-
pany called IHomer a significant problem is being discussed.
Because the default work location in the company is home
the people work distributed from each other most of the time,
and have always physically met up once a week on Tuesdays
in order stay connected. However, one employee has been
away on contract for a prolonged period now, preventing him
from attending the weekly meet-ups and he is starting to feel
more and more disconnected from the rest of the team. On
the one hand he is starting to lose touch with what occupies
his colleagues and on the other hand his colleagues weren’t
even aware he was feeling unhappy with the overall situation.
Fast forward almost two years to the present day and the
issue has been dealt with by the introduction of WeHomer, a
microblogging environment in which people at IHomer share
their activities and moods with each other, to stay current on
each other’s feelings, experiences and latest exploits and as a
result stay connected as a team.
The story above is not unique as Software Engineering
is becoming more and more distributed. This is caused by
the increasing globalization of business [1], [2], [3] and the
rising popularity of working from home [4]. At the same
time, significant challenges are faced when collaborating in
a distributed setting as reported in the well-known work of
Olson and Olson [5]. More recently Nguyen et al. [6] have
investigated whether the effects of distance on distributed
communication delay and task completion reported in the
literature still exist. They report that advances have been made
but also that more work is needed and there are still many open
research questions.
We believe both microblogging and mood sharing are
essential to alleviate challenges arising from this distributed
nature. On the one hand microblogging is essential, since
being able to exchange small elements of content makes
people feel more connected with others, especially when
people work distributed from their colleagues [7], [8], [9].
On the other hand mood sharing is essential, since being
aware of the emotional state of your colleagues makes it
possible to act accordingly and achieve better results in
joint work [10]. Therefore, in this research, we aim to
determine whether an environment in which one can both
express himself and get a sense of how his team members
feel is valuable to distributed software engineers. In this
paper, we use a microblogging solution extended with
mood indicators (MBMI) called WeHomer to learn from
the use of such an environment. The main goal of the paper is:
”To understand how microblogging with mood-indicators
helps distributed organizations in knowledge sharing”
Furthermore we have identified four research questions:
What sort of topics are discussed in MBMI?
What is the impact of the introduction of a MBMI on a
software team?
What is the impact of distribution on the use of a MBMI?
How does team composition impact collaboration with
We structure this paper as follows: In the next section we
discuss related work to this research. In section III we discuss
the research site and methods of data collection and analysis.
Subsequently we show descriptive statistics in section IV and
present the most important findings in section V. Finally we
discuss threats to the validity of our study in section VI and
conclude upon our work and discuss future work in section
Software Engineering is by nature a highly collaborative
activity, and having access to knowledge about the context in
which you are working is essential to properly collaborate with
others [11], [12]. In literature this kind of knowledge is com-
monly referred to as ‘awareness‘ [13], [11]. When working
co-located this information is exchanged relatively passively
and unobtrusively [11], [14], so people are continuously aware
of information related to their current context [15]. However,
when people no longer share a physical work environment
exchanging awareness information without technological sup-
port becomes unfeasible [15]. Therefore, the (Global) Software
Engineering community has developed and studied a wide
variety of tools, for example: Instant Messaging, email, issue
management systems and configuration management systems.
According to Bly et al. [16] it is particularly important
to recognize the need for informal interactions, spontaneous
conversations, and general awareness of people and events
when teams are geographically dispersed. Storey et al. ad-
vocate further research on understanding how social media
plays a role in (Global) Software Engineering [17]. One of
the potential implications they identified in their research,
concerns the challenges which arise when teams are distributed
across time zones and geographical locations, and lack infor-
mal mechanisms for communication. They emphasize social
media is regarded as a mechanism that can support informal
and serendipitous interactions across the team, and as such can
alleviate these challenges. Finally, they present ten research
questions, of which the following three are applicable to this
1) Can social media play an effective role in supporting
coordination and task articulation?
2) What kinds of social media would increase informal
communication, the flow of knowledge and awareness
across team and project members?
3) What are the drawbacks from increased transparency in
team projects? Does this lead to privacy concerns?
There are several social media tools available to software en-
gineers which facilitate coordination, communication with and
learning from other users, being informed about new devel-
opments and creation of informal documentation [17]. These
tools can be characterized by an underlying ‘architecture of
participation‘; systems that are designed for user contribution
[18]. Such a design supports the creation of collective value,
often as an automatic byproduct of an individual activity.
Wikis, blogs and microblogs are some well-known examples
of such social media solutions.
In our research we aim to determine whether an environment
in which one can both express himself and get a sense of
how his team members feel is valuable to distributed Soft-
ware Engineering teams. The specific environment we used,
called WeHomer, allows users to exchange small elements
of content such as sentences, images and hyperlinks, and
their corresponding emotion. In fact, we use a microblogging
solution extended with mood sharing functionality. Several
research projects have been conducted in the field of Software
Engineering to increase the understanding of how and why
people use microblogging solutions. We will consider three
of these user studies, namely studies on Twitter [7], Yammer
[8] and BlueTwit [9], to identify similarities and differences
between these microblogging solutions and WeHomer.
Firstly, Twitter is a publicly available microblogging service
with which users can publicly share messages, limited to 140
characters. They can also indicate whether their messages are
public or private. When messages are indicated to be public
they are accessible to all users of Twitter. However, when
messages are indicated to be private they are only accessible to
those users who have subscribed and are explicitly authorized
to the user his feed. Zhao et al. [7] identified several char-
acteristics in the use of Twitter. Important examples are: (i)
frequent small updates of personal life events enable users to
stay aware of people they do not encounter on a daily basis and
(ii) subscribing to people you personally know and selected
enables users to get trustworthy and useful information.
Secondly, Yammer1is very similar to Twitter, the main dif-
ference being that Twitter is publicly available while Yammer
is enclosed by organizational boundaries. Other differences are
the absence of a character limit, the possibility to create private
and public groups and the opportunity to add attachments to
messages. Zhang et al. [8] also identified several characteristics
of the use of Yammer. They give an indication that users use
Yammer more for publishing news about their groups or busi-
ness units than for news about themselves. Subsequently, they
indicated that Yammer was used to have long conversations
and discussions. Finally, they found that Yammer enables users
to stay aware about what others are working on and to make
new connections.
BlueTwit is also very similar to Twitter, however it differs on
two points: (i) BlueTwit is only available within organizational
boundaries and (ii) BlueTwit has a character limit of 250
instead of the 140 character limit of Twitter [9]. Ehrlich et
al. [9] discussed characteristics of the use of BluetTwit, It
enables: (i) having internal conversations about confidential
information, (ii) staying aware of what others are working on
and (iii) enhancing your reputation.
All of the above user studies on the usage of different
microblogging solutions mentioned an important side effect
of microblogging in general: people feel more connected
with each other. This is especially the case when people
work distributed from their colleagues since microblogging
kept them connected to other colleagues and the company,
and alleviated the feeling of isolation. WeHomer differs from
BlueTwit in the sense that it drops the character limit, enforces
subscription to all users automatically (manageable because of
the small community), and adds mood sharing functionality.
Garcia et al. [10] introduced ‘Emotional Awareness‘ and
argue that it enables users to become aware of the emotional
state of their collaborators and act accordingly to achieve better
results in their joint work. Other research mainly focuses on
electronic meeting systems in which each participant explicitly
specifies his mood and changes in the average mood are
visualized to all participants [19], [20]. WeHomer integrates
both microblogging functionality and the opportunity to ex-
press your current mood into a single environment. This is
the main differentiator of WeHomer in comparison with other
microblogging solutions.
A. Research Site
This research is carried out at IHomer, a Dutch Software
Engineering company founded in August of 2008 in which
it is common practice to work from home. IHomer currently
employs 20 people, working on a variety of products, projects
and contracts. The largest team consists of 7 people work-
ing on related projects, but the overall group is very close
with personnel moving between teams and teams exchanging
projects as needed. Even though it is common practice in the
company to work from home, the employees try to get together
once a week on Tuesdays to meet face-to-face at an office
to stay connected. Sometimes this can be difficult however,
for example when someone is away on a contract and has
other obligations on Tuesdays. The company has grown over
the past years and initially on Tuesdays everyone discussed
what they were doing. This worked well until the company
size reached 16, and then sub teams were formed to keep
this face-to-face communication more tractable. Teams cluster
according to various factors: projects and related technologies
being two of them.
People at IHomer aim to work together closely and stay a
very connected community. One of the core strategies to stay
connected is the weekly face-to-face meeting. As mentioned
above this is not always feasible which can become a practical
problem if people are unable to attend the majority of the
Tuesday meet-ups for a prolonged period. In order to cope
with these issues the WeHomer system was developed by an
employee of the company (not an author of this paper) and
deployed in October 2011. It is a platform on which IHomers
can share information about their day with their colleagues
in order to stay connected and increase awareness. Users can
share information about a new topic, called an entry or respond
to an existing topic, called a comment (commenting was not
supported in the first three months of our data analysis period).
Comments are shown in chronological order grouped together
under the entry to which they correspond. We use the term
post to refer to something that is either an entry or a comment.
Posts cover such items as what you are doing right now, what
you have done, what you are going to do, how you feel about
something and random thoughts.
Associated with each post is a happiness score ranging
from 0 (totally unhappy) through 100 (utter bliss) depicting
how the user feels about this post. In the user interface of
WeHomer (see figure 1) the happiness index can be selected
by use of a slider bar which shows one of 5 discrete emoticons
corresponding to the level that is selected by the user2. The
exact integer value of the happiness index is not derivable by
end users.
2Happiness index ranges: ’>:-(’ = [0,20), ’:-(’ = [20, 40), ’:-|’ = [40, 60],
’:-)’ = (60, 80], ’:-D’ = (80, 100]
Fig. 1. WeHomer user interface
B. Method
The primary method of data collection we use is mining the
WeHomer data between October 2011 and November 2012 by
analyzing and subsequently coding the content. During this
period there were a total of 1312 entries and 1189 comments.
Because it is feasible to hand code each of these entries and
comments sampling was unnecessary and we analyzed the
content of all entries and comments. The coding was done by
the first two authors and the coding set was arrived at in an
iterative fashion. Firstly, a random (but consecutive) sample
of 50 entries (including the corresponding comments) was
selected and coded independently by both coders. Following
this the two coders compared their codes and discussed
their reasoning behind those codes. Based on this discussion
the coders agreed on a joint coding set with which they
independently coded another random (consecutive) sample of
25 entries (again including the corresponding comments) and
discussed discrepancies in how they coded the sample. Based
on this discussion they refined the coding set and did another
iteration. After a total of three such iterations they decided
the coding set was consistent between them and they could
go ahead with the actual coding. To do this they divided the
total data set in six ranges of approximately 200 entries and
each coded three non-consecutive ranges.
Subsequently, based on what we found in the content
analysis we conducted semi-structured interviews3with five
of the nineteen users in which we asked questions about what
was unclear to us in the analysis and follow-up questions
we had based on the analysis. To select which five people
to interview we used purposive sampling in order to get an
as complete view as possible. In the selection process we
explicitly excluded authors of this paper. We did select the
3Interview structure can be found at:
original developer of the WeHomer system, the person with
the highest number of posts, the one with the lowest number
of posts and the two people with respectively the highest and
lowest number of entries to number of posts ratios.
In this section we present information derived from the
mined data to present the reader with an image of how the
MBMI environment is used.
In figure 2 the weekly average number of entries, comments
and posts is shown. In this figure it can there is significant and
consistent use of the WeHomer system for the entire year we
are investigating. Additionally we can see that at the start of
the period there were no comments and considerably more
entries than the remainder of the period. This is because at
the start posting comments was not possible and people used
entries to comment on another entry by referring to the entry
they wished to comment on.
Fig. 2. The number of entries, comments and total posts per week
Subsequently, in table I we present the median length and
the interquantile range of entries, comments and for entries and
comments combined to give an indication of their respective
lengths. So, for instance, we see that 75% of the entries are
shorter than 143 characters.
Lower Quartile Median Upper Quartile
Entries 62 97 143
Comments 22 49 86
Posts 39 75 122
In figure 3 for the entire year the average happiness per
week as well as the maximum and minimum happiness score
are shown. In this figure it can be seen that the happiness
fluctuates significantly and that in general the highest and
lowest score for a week lie relatively far apart.
Fig. 3. The average, lowest and highest happiness score per week
Further, in WeHomer a default happiness score is automat-
ically selected for each post a user makes, namely 70 on the
range between 0 and 100. If a user doesn’t manually select
another happiness score this default score is used. Therefore it
is interesting to investigate how often the users deviated from
this default value. In table II we show separately for entries
and comments how often this occurred. In this table we can
see that a significant portion of the happiness scores were
set at the default score. Therefore, we asked the five people
we interviewed whether these values were chosen consciously.
All the interviewees told us that even though they on occasion
forgot to change the happiness score, in general they spent a
minute to think which score to select, even if this is the default
happiness score.
Default Non-Default Total
Entries 791 60.3% 521 39.7% 1312 52.5%
Comments 931 78.3% 258 21.7% 1189 47.5%
Total 1722 68.9% 779 31.1% 2501 100%
Finally, in IHomer there exist 4 teams of people working
on related projects. We compare the amount of directed
communication between members of the same team and
members of different teams to give an indication of how much
collaboration was being done overarching the different teams.
To do this we do the following: Firstly, we define commenting
on an entry posted by a specific user as the utilization of a
directed communication line between those two users. We do
this even though all other users can see this communication
because the communication is at least directed at that specific
user. Subsequently we sum all of these utilizations of commu-
nication lines. An interactive visualization of the utilization of
communication lines between both team members and non-
team members can be seen at
Following this we needed to compare the amount of com-
munication within teams with the amount of communication
outside of teams. We did this by doing the following for each
1) Count the total number of comments of the people in
the current team on other people within the current team
2) Count the total number of comments of the people in the
current team with the people outside the current team
3) Make the number found in step 1 relative to the team
size (n) by dividing it by n*(n-1) (n*(n-1) to represent
the total number of communication lines)
4) Make the number found in step 2 relative to the number
of people outside of the team
A summary of these results is shown in table III:
Team Step 1 Step 2 Step 3 Step 4
Team 1 (5 people) 49 162 2.45 0.89
Team 2 (7 people) 386 194 9.19 1.47
Team 3 (5 people) 47 161 2.35 0.88
Team 4 (2 people) 24 165 12.00 0.61
In this table it can be seen that the utilization of com-
munication lines within teams is considerably higher than
communication lines crossing team boundaries.
In this section we answer the research questions specified
in section I. We structure the section based on the research
questions, answering each research question in a separate
A. Topics
Research question 1 is: ”What sort of topics are discussed in
MBMI?” We answer this research question by first discussing
the set of codings we used to code the data to give insight
in the variety of topics discussed in the MBMI system.
Subsequently we show the occurrence of each of the codings in
both the entries and comments to give insight in the frequency
each topic occurred. Finally we analyze these occurrences and
make generalizations.
As described in section III we used an iterative bootstrap-
ping process to construct the coding set. This coding set is
structured in four major categories:
1) Nature
2) Form
3) Intention
4) Content
Each of these coding categories is further divided into
sub-categories and actual codes. In appendix A we show
this subdivision for each of the four major coding categories
depicted as trees. In these trees leaf nodes depict actual codes
while non-leaf nodes depict sub-categories. In the appendix
we also explain each of the codes and give examples.
Subsequently we applied all four coding categories4to the
entries and the first two coding categories to the comments. We
only apply codes of the first two categories to the comments
because comments are made in the context of an entry which
makes it difficult to differentiate in how far the intention and
content of a comment are dictated by the corresponding entry.
We present the occurrence of each of the codes in the set
of entries and the set of comments in figure 4 and figure 5
Fig. 4. The frequencies of the codings for the entries
Fig. 5. The frequencies of the codings for the comments
4Note that the codes in the four categories are not mutually exclusive in
their application to posts (e.g. a post can contain a positive and negative part)
In figure 4 it can be seen the most frequently occurring
codes are ’Statement’ (85.5%), ’Coordination’ (73.9%), ’Pos-
itive’ (58.5%), ’Work Planning’ (54.9%) and ’Personal Infor-
mation’ (49.8%). In addition to this, four of the five people
we interviewed indicated they have the tendency to post more
positive things and to post more when they are in a positive
mood. Further the consensus in the interviews about what they
post is that it is everything they consider useful or interesting
to their team members. One would suspect this to lead to a
diverse list of topics to be shared on the medium and both the
diversity of codes and the relatively distributed occurrence of
these codes support this expectation. Finally, when asked to
specify what they shared most, popular subjects mentioned are:
personal information, project information with the intention
to coordinate, technical information and prospects. This also
corresponds with the actual data presented in figure 4.
Subsequently we discuss a deviation from the expected. An
evident application for a MBMI system is asking questions.
However, we found a relatively low amount of these. The total
number of entries that are questions is 9.1% and the total
number of comments that are questions is 7.5%. We compared
these numbers to the results found in a study by Erhlich et al.
[9] on the investigation of the usage of Twitter and BlueTwit
(an internal proprietary version of Twitter) in an organization
for people that use both tools. To compare our number of
questions to theirs we summarized the percentages for ”Ask
Question” and ”Directed with Question” in their result set to
yield a total number of questions of 6% for Twitter and 13%
for BlueTwit (versus our 7.5% and 9.1% respectively). So, in
the environment in our study relatively more questions were
asked than in the Ehrlich setting for Twitter use, but less than
with BlueTwit.
When asked about the amount of questions in the five
interviews the respondents indicated that they asked a rela-
tively low amount of questions on the MBMI system for two
reasons. Firstly, the medium is asynchronous which makes it
unpredictable when a question will be answered. Secondly,
they indicated they usually knew who to contact (or at least
knew who knew who to contact) and preferred contacting
someone who would know the answer directly over asking it
to the entire group. It was also discussed in the interviews the
low amount of questions is likely to be specific to companies
with a relatively small size, close personal connections and
transparency between the team members because in such
companies people will more easily know who knows what.
B. Impact on a Software Team
Research question 2 is: ”What is the impact of the introduc-
tion of a MBMI on a software team?” The first main impact we
found is that members of software teams feel more connected
to each other when they are able to share activities and moods.
We base this primarily on information from the interviews.
Firstly, three out of the five people we interviewed explicitly
reported feeling more connected to their colleagues since they
were able to share their moods and activities within the team
using the MBMI system. Additionally these interviewees also
reported being better able to understand their colleagues since
using the environment and two of them reported they felt
their colleagues understood them better as well. This finding
corresponds to the results of the studies of Zhao et al. [7] and
Ehrlich et al. [9] on microblogging in the work place. Zhao
et al. [7] conclude: ”Our results suggest that microblogging
may help colleagues to know each other better as persons,
that is in addition to professional relationships; this benefit
is achieved by staying aware of small details about others’
personal lives, interests, and current moods, which in turn
creates more opportunities for exchanging acknowledgments
and social support, generating new common ground, and
creating and sustaining a feeling of connectedness.
Lesson Learned 1
Distributed software engineers feel more connected to each other
when they are able to share activities and mood
As the second main finding, we found a MBMI makes
information that is traditionally harder to consistently acquire
more approachable and less volatile. This is based on both the
content analysis and the interviews. In the content analysis we
found that in particular the coding-categories entrepreneurial
tasks (14.5% of all entries) and customer relations (9.3% of
all entries) represent a considerable portion of the data. One
of the interviewees indicated information about these types of
activities is traditionally difficult to consistently gather. For
instance information about ”how to build a business” is often
shared in face-to-face communication which makes it difficult
to acquire at a later time (you will have to ask or try to recall)
and the information is likely to be different from the original.
In the interviews people comment they consider it a strength
of the system to be able to share non-time critical information:
Information they would like to know about ”eventually, but not
necessarily within the next five minutes”. Before WeHomer
communicating this type of information was often postponed
until a weekly face-to-face meeting or discarded altogether.
One of the interviewees even said he found the system to
offer benefits over meeting face-to-face on Tuesdays: ”With
WeHomer it is easier to stay current than by meeting people
face-to-face on Tuesdays because then you don’t get to talk to
Lesson Learned 2
A mood-activity environment makes information that is traditionally
harder to consistently acquire more approachable and less volatile
Finally, we also found a MBMI system facilitates an unob-
trusive way to express your personal feelings or thoughts to
your colleagues. All of the interviewees mentioned they con-
sidered the low threshold the environment offered for sharing
information with their colleagues an important strength. One
of the interviewees said he felt he could ”share knowledge
and emotions like you are co-located”. We can also see the
environment offers a light-weight method to share personal
information since over half of the entries (52%) contain
personal information.
C. Impact on a Distributed Software Team
Research question 3 is: ”What is the impact of distribution
on the use of a MBMI?” Firstly, we found that people who
work co-located with the majority of their team, share less
activities and moods with those team members that are non-
colocated. In our setting this behaviour presented itself as
follows: while for the rest of the week the default work
location is home, on Tuesdays people at IHomer try to work
co-located at a central office as much as possible. However, at
times this is unfeasible for specific team members, for instance
due to being contracted at a customer location. In practice at
least half of the team is present on Tuesdays the vast majority
of the time, but it is rare for the entire team to be present.
Therefore it is striking to see that on Tuesdays the number
of entries and comments is significantly lower than on other
days of the week (see figure 6). Finally, also in the interviews
it was recognized that ”on Tuesdays WeHomer is used very
little”. This is similar to what was reported in [21] on the
deployment of a conversation overhearing tool in an industrial
setting with a non-homogeneous geographical distribution.
Fig. 6. The number of entries and comments for each day of the week
Lesson Learned 3
In distributed Software Engineering teams, people who work co-
located with the majority of their team, share less activities and
moods with those team members that are non-colocated.
Paradoxically to what is discussed above, the interviewees
indicated they do find it particularly valuable to share moods
and activities with their distributed colleagues. One of the
interviewees stated: ”WeHomer is used very little when people
work co-located on Tuesdays which makes things less trans-
parent for people that cannot be there”. The interviewees
indicated they recognize the value in making sure the entire
team stays connected, even when part of the team works co-
located and part of the team works distributed. They recognize
the value because they know from experience how difficult
being the dislocated colleague can be. It is striking to see that
even though they know it is important to help their distributed
team members, they still struggle to do so.
Lesson Learned 4
In distributed Software Engineering teams, people who work co-
located with the majority of their team, find it is particularly useful
to share activities and moods with those team members that are non-
A useful insight on this is also shared by Fullerton of
Stack Exchange in his blog post [22] about the lessons
learned from three years of working in a distributed team.
He states: ”There’s no halfsies in a distributed team. If even
one person on the team is remote, every single person has to
start communicating online. The locus of control and decision
making must be outside of the office: no more dropping in
to someone’s office to chat, no more rounding people up to
make a decision. All of that has to be done online even if the
remote person isn’t around. Otherwise you’ll slowly choke off
the remote person from any real input on decisions.
D. Impact of Team Composition on MBMI
Research question 4 is: ”How does team composition impact
collaboration with MBMIs?” Our main finding with respect
to this research question concerns the regularity in which
the members of a team use the MBMI: do all the members
of the team use the MBMI system an equal amount of the
time and for the same sort of topics? If this usage differs
significantly between different team members the distribution
of relevant information in the team will be unbalanced as
well. This was the main challenge in the use of the MBMI
system that came forward in the interviews. Examples of
things interviewees said are: ”The success of WeHomer is
dependent on participation”,”There is only a challenge for
those not using it” and ”If you don’t participate you miss
things”. The challenge is threefold. Firstly, team members
using the environment less than their colleagues run the risk of
missing things. Secondly, a team member sharing information
cannot be certain his team members know about this. Finally,
since the MBMI system isn’t used for everything, you cannot
infer something did not happen because it is not available
there. Therefore, to have a complete view of all valuable
information about the mood and activities of team members
software engineers need to consult other sources as well.
Lesson Learned 5
If the regularity with which team members utilize the mood-activity
environment differs, the distribution of relevant information in the
team will be unbalanced
Finally, in the interviews we talked about the type of
teams for which they thought a MBMI system would be
beneficial. Firstly, to use the system in the same (personal)
fashion as it is being used at IHomer the teams need to be
sufficiently involved. One of the interviewees said: ”I need
to know the people” while another said that ”teams need to
be homogeneous (shared interests, shared work)”. They also
mentioned that as a direct result of this, team size can become
an issue but only of it leads to the teams members being less
involved with each other.
On the other side of things the interviewees did consider
a MBMI system beneficial to all companies. One interviewee
said: ”every organization needs a WeHomer because even a
closed door is a barrier”. They do believe however that the
type of information that is being shared will be connected to
the type of organization. For instance one of the interviewees
explained that the he considered the high amount of personal
information and the diversity of the messages on WeHomer
to be tied to the open character of IHomer and that he would
expect a more traditional organization to share a larger portion
of technical information instead.
Threats to external validity can exist at each of the levels of
generalization in a study. In our study, a threat to external
validity exists in the generalization of the single Software
Engineering team to the population of all distributed Software
Engineering teams. To be able to better generalize beyond the
setting we performed the study in, the study should be repeated
in other teams as well. With respect to the generalization of
the sampled data to the population of IHomer our work is
much less threatened. For the interviews we sampled 5 out of
the 20 people in the team (25%) and for the content analysis
we even coded 100% of the posts in WeHomer for the year
of data we investigated.
Furthermore, there exist threats to construct validity in our
study. Firstly, we attempted to mitigate threats to reliability
by elaborately describing our research site and methods and
making our coding set and interview design available. Next
to this we also make all of our data available in anonymized
form and make the tool available upon request. We do this
to make both our data gathering methods and the analysis
of our data, repeatable. Subsequently, a threat to construct
validity is mono-operation bias. Because we only researched
the application of MBMI environments with one specific tool,
one could argue the results only apply to the use of that
tool. The only mitigation we need to offer for this is the
general nature of the tool itself. Basically any tool with which
it is possible to share activities and moods in distributed
teams will suffice and the WeHomer tool clearly fulfills these
requirements. A final threat to construct validity in this study
is that both the creation of the coding set and the coding of
the posts was done by the first two authors who are also
employees in the company at which the study was being
performed. The advantage of this is that the researchers posses
insight knowledge and can leverage this to code the data
more accurately. A disadvantage is that the researchers might
not be completely impartial due to their involvement in the
setting. Overall, it is our opinion the advantages outweigh the
The main contributions of this paper are the answers to
the research questions. First, we answered what sort of topics
are discussed in microblogging systems with mood indicators
(MBMI) by presenting the nature, form, intention and content
of the posts in over one year of usage data and presenting
the frequency at which these occurred. Based on this data
we found that distributed software engineers primarily share
positive posts with the intention to either coordinate or provide
personal information. Furthermore, when compared to other
corporate microblogging solutions we found a relatively low
amount of questions.
Subsequently, we have shown there are two major impacts
of the introduction of MBMI on a software team. Firstly,
team members become more connected. The loss of teamness
is a major and unresolved issue in the field of GSE and
therefore advances in this area are significant. Secondly, the
MBMI system made information that is traditionally harder to
consistently access more approachable and less volatile
Further, on the impact of distribution of the software team
on the use of MBMI, we found that the way people act when
working co-located with the majority of their team is para-
doxical to how they think they should act. On the one hand,
people share less activities and moods with their distributed
colleagues while on the other hand they do recognize the value
in staying connected with the rest of the team. It is striking to
see that even though they know it is important to help their
distributed team members, they still struggle to do so.
With respect to the fourth research question, on the impact
of team composition on collaboration with MBMI, we found
that the distribution of relevant information in the team will be
unbalanced if team members use the environment unequally.
Concerning future work we are particularly interested in
researching the actual value of incorporating mood in mi-
croblogging systems. A way to accomplish this is to perform
two user studies in similar software teams in which one of the
teams receives access to a regular microblogging solution and
the other team receives access to a system that is similar in
every way except the addition of the possibility to share mood
with team members.
[1] E. Carmel, Global software teams: collaborating across borders and
time zones. Upper Saddle River: Prentice Hall PTR, 1999.
[2] J. Herbsleb and D. Moitra, “Guest Editors’ Introduction: Global Soft-
ware Development,IEEE Software, vol. 18, no. 2, pp. 16–20, 2001.
[3] J. Herbsleb, “Global Software Engineering: The Future of Socio-
technical Coordination,” in Proceedings of the IEEE 2007 Workshop
on the Future of Software Engineering. IEEE Computer Society Press,
2007, pp. 188–198.
[4] The Dieringer Research Group Inc., “Telework Trendlines 2009: A
Survey Brief by WorldatWork,” 2009.
[5] G. Olson and J. Olson, “Distance matters,” Human-computer interaction,
vol. 15, no. 2, p. 139178, 2000.
[6] T. Nguyen, T. Wolf, and D. Damian, “Global software development and
delay: Does distance still matter?” in Proceedings of the IEEE 2008
conference on Global Software Engineering. IEEE, 2008, pp. 45–54.
[7] D. Zhao and M. B. Rosson, “How and why people twitter: the role
that micro-blogging plays in informal communication at work,” in
Proceedings of the 2010 International Conference on Weblogs and
Social Media. ACM Press, 2009.
[8] J. Zhang, Y. Qu, J. Cody, and Y. Wu, “A case study of micro-blogging
in the enterprise: use, value, and related issues,” in Proceedings of the
28th international conference on Human factors in computing systems.
ACM, 2010, pp. 123–132.
[9] K. Ehrlich and N. S. Shami, “Microblogging inside and outside the
workplace,” in Proceedings of the ACM 2009 international conference
on Supporting group work. The AAAI Press, 2010.
[10] O. Garc´
ıa, J. Favela, and R. Machorro, “Emotional awareness in
collaborative systems,” in String Processing and Information Retrieval
Symposium, 1999 and International Workshop on Groupware. IEEE,
1999, pp. 296–303.
[11] K. Schmidt, “The Problem with ‘Awareness’: Introductory Remarks on
‘Awareness in CSCW’,” Computer Supported Cooperative Work, vol. 11,
no. 3-4, pp. 285 – 298, 2002.
[12] A. Syri, “Tailoring cooperation support through mediators,” in Pro-
ceedings of the 1997 European Conference on Computer Supported
Cooperative Work. Kluwer Academic Publishers, 1997, pp. 157–172.
[13] P. Dourish and V. Bellotti, “Awareness and Coordination in Shared
Workspaces,” in Proceedings of the ACM 1992 Conference on Computer
Supported Cooperative Work. ACM Press, 1992, pp. 107–114.
[14] J. Fogarty, S. Hudson, C. Atkeson, D. Avrahami, J. Forlizzi, S. Kiesler,
J. Lee, and J. Yang, “Predicting human interruptibility with sensors,”
ACM Transactions on Computer-Human Interaction, vol. 12, no. 1, pp.
119–146, 2005.
[15] K. Dullemond, B. van Gameren, and R. van Solingen, “Virtual open
conversation spaces: Towards improved awareness in a GSE setting,” in
Proceedings of the 2010 International Conference on Global Software
Engineering. IEEE Computer Society Press, 2010, pp. 247–256.
[16] S. A. Bly, S. R. Harrison, and S. Irwin, “Media spaces: bringing people
together in a video, audio, and computing environment,Commun.
ACM, vol. 36, no. 1, pp. 28–46, Jan. 1993. [Online]. Available:
[17] M.-A. Storey, C. Treude, A. van Deursen, and L.-T. Cheng, “The impact
of social media on software engineering practices and tools,” in Pro-
ceedings of the FSE/SDP workshop on Future of software engineering
research. ACM, 2010, pp. 359–364.
[18] T. O’Reilly, “The architecture of participation,” http://www.oreillynet.
[19] S. Mora, V. Rivera-Pelayo, and L. M ¨
uller, “Supporting mood awareness
in collaborative settings,” in Proceedings of the 2011 International
Conference on Collaborative Computing: Networking, Applications and
Worksharing. IEEE, 2011, pp. 268–277.
[20] A. Fessl, V. Rivera-Pelayo, V. Pammer, and S. Braun, “Mood tracking
in virtual meetings,” in Proceedings of the 2012 European Conference
of Technology Enhanced Learning. Springer, 2012, pp. 377–382.
[21] K. Dullemond and B. van Gameren, “An industrial evaluation of
technological support for overhearing conversations in global software
engineering (in press),” in Proceedings of the 2012 International Confer-
ence on Global Software Engineering. IEEE Computer Society Press,
[22] D. Fullerton, “Why we (still) believe in working remotely,” http://blog. we-still- believe-in-working-remotely/,
The used coding set is the following. Firstly we defined four
major types of codes.
1) Nature
2) Form
3) Intention
4) Content
Each of these coding categories is further divided into
sub-categories and actual codes. Below we will show this
subdivision for each of the four major coding categories. In
these trees leaf nodes depict actual codes while non-leaf nodes
depict sub-categories. Below each figure we will explain the
codes and give examples. Firstly for all entries and comments
we coded the nature of the message (see figure 7). This depicts
whether the content of the post can be considered positive,
negative or neutral. So, for example a post stating a new
assignment has landed is positive while a post about a failed
build or a sick family member is negative.
Fig. 7. The nature of an entry or comment
Secondly we also coded the form of the entries and com-
ments (see figure 8). With the codings we mean the following:
Statement - An assertion
Answer - Attempt to answer a question asked in a
previous post
Joke - Something said or done to provoke laughter
or cause amusement
Compliment - Expression of praise, commendation,
or admiration
Best Wishes - Wishing something nice to someone
(such as: ”good luck” or ”happy birthday”)
Standard - A statement, but not an answer, joke,
compliment or best wishes
Question - Attempt to illicit an answer to some question
Best Wishes
Fig. 8. The form of an entry or comment
We choose to only apply the final two coding categories
(intention and content) to the entries and not the comments.
We elect to do so because comments exists only in the context
of the entry to which they belong. Because of this, comments
are often quite brief and leave out much of this contextual
information making it is infeasible for coders to consistently
decide what intention and content is specifically applicable to
that comment. The types of intention we distinguish are de-
picted in figure 9. With these codings we mean the following:
Sharing personal information - Intention to share in-
formation that is about the poster and his/her personal
Sharing work related information - Intention to share
information that is about the work of the poster
Coordination information - When the intention is
to coordinate with colleagues
Knowledge - when the intention is to share factual
Social interaction - Intention to follow social protocol
for making and maintaining relationships with others
Fig. 9. The intention of an entry
Finally the content coding is shown in figure 10. With these
codings we mean the following:
Information about a person
Health - The poster’s health
Sentiment - How the poster feels
Personal Experience - Information about an expe-
rience which is not primarily work-related (Travel,
Family, Scenario)
Information about technology
Technical Knowledge - Specialist information
Information about task articulation work - Information
about work which is done to support the core activities
of IHomer (including infrastructure and planning)
Work Planning - When work will be done or is
– Work Assignment - Who will do certain work
(Assignment, Expertise Finding)
Supplies - Information about supplies (Including for
instance food)
– Non-Technical Infrastructure - For instance the
office or office equipment
Technical Infrastructure Intern - For instance
IHomer’s timesheet application
– Technical Infrastructure Extern - For instance
DNS, Skype, IDE and phoneline
Information about customer relations - Information
about relations with customers and the process around
– Relation - Directly relating to the making and
maintaining of the professional relationships with
business relations
– Project Commissioning - About transferring fin-
ished (or partially finished work) to the customer
(including things such as training)
Information about entrepreneurial tasks - Tasks di-
rectly related to the organization, operation and manage-
ment of risk with respect to a business venture
Prospects - Opportunities for new work or projects
Company Meeting
Applicants - Hiring new people to work for IHomer
Invoicing - About actions needed to get paid by the
customers (such as sending the actual bill)
Information about
a person
Information about
Information about
task articulation
Work Planning
Work Assignment
Information about
Information about
Fig. 10. The content of an entry
... On the other hand, the relevance of moods and emotions at work has been confirmed by several studies on workplace communication. Dullemond et al. [2013] investigated mood sharing in a micro-blogging system to facilitate knowledge sharing within a fully distributed team of software engineers and showed that members of software development teams feel more connected to each other when they are able to share activities and moods. Although sharing moods and investigating team-connectedness are points in common with our work, our research differs in considering the benefits of self-tracking on individual work performance as well as the impact at organizational level. ...
... Work done in HCI has confirmed the importance of emotions not only for self-tracking purposes in private areas of life but also in work settings to better support the reevaluation of past working situations based on emotions [McDuff et al. 2012], as well as to improve communication in work environments [De Choudhury and Counts 2013;Dullemond et al. 2013]. However, interviews in our study revealed that due to time and business constraints, time and space for reflection by reviewing their gathered mood data was not given to call takers. ...
The benefits of self-tracking have been thoroughly investigated in private areas of life, like health or sustainable living, but less attention has been given to the impact and benefits of self-tracking in work-related settings. Through two field studies, we introduced and evaluated a mood self-tracking application in two call centers to investigate the role of mood self-tracking at work, as well as its impact on individuals and teams. Our studies indicate that mood self-tracking is accepted and can improve performance if the application is well integrated into the work processes and matches the management style. The results show that (i) capturing moods and explicitly relating them to work tasks facilitated reflection, (ii) mood self-tracking increased emotional awareness and this improved cohesion within teams, and (iii) proactive reactions by managers to trends and changes in team members' mood were key for acceptance of reflection and correlated with measured improvements in work performance. These findings help to better understand the role and potential of self-tracking at the workplace, and further provide insights that guide future researchers and practitioners to design and introduce these tools in a work setting.
... Recently, several researchers investigated and evaluated the the role played by communications in Twitter and more in general, the role played by "microblogging", in software development organizations [34], [35], [36], [37]. Zhao et al. [37] surveyed 11 microblog participants to better understand the conversational aspects of Twitter discovering the potential benefits it brings to informal communication at work. ...
... Moreover, Ehrlich et al. [34] showed how different the use of the external/internal microblogs are: external microblogs are used for sharing general information; instead, internal microblogs are used to technical assistance and discussion. Finally, Dullemond et al. [35] evaluated microblogging discussions, and found how "mood-activity environment" helps to obtain information that is traditionally harder to obtain in a less volatile form. In summary, although there are barriers, microblogging could likely become another promising communication channel. ...
Conference Paper
Full-text available
Written communications recorded through chan-nels such as mailing lists or issue trackers, but also code co-changes, have been used to identify emerging collaborations in software projects. Also, such data has been used to identify the relation between developers' roles in communication networks and source code changes, or to identify mentors aiding newcomers to evolve the software project. However, results of such analyses may be different depending on the communication channel being mined. This paper investigates how collaboration links vary and complement each other when they are identified through data from three different kinds of communication channels, i.e., mailing lists, issue trackers, and IRC chat logs. Also, the study investigates how such links overlap with links mined from code changes, and how the use of different sources would influence (i) the identification of project mentors, and (ii) the presence of a correlation between the social role of a developer and her changes. Results of a study conducted on seven open source projects indicate that the overlap of communication links between the various sources is relatively low, and that the application of networks obtained from different sources may lead to different results.
... They found that employees use Yammer for news about groups and less for posting content about themselves. Dullemond et al. [19] also studied the use of a microblogging tool within a small distributed software company where moods are communicated explicitly as part of posts. ...
... Sites such as GitHub provide rich opportunities for use in university courses-Arie van Deursen discusses how he successfully used GitHub in a software architecture course 19 . Furthermore, online learning environments, such as Khan Academy and Coursera, provide additional mechanisms to foster self-paced learning in online communities. ...
Conference Paper
Full-text available
Software developers rely on media to communicate, learn, collaborate, and coordinate with others. Recently, social media has dramatically changed the landscape of software engineering, challenging some old assumptions about how developers learn and work with one another. We see the rise of the social programmer who actively participates in online communities and openly contributes to the creation of a large body of crowdsourced socio-technical content. In this paper, we examine the past, present, and future roles of social media in software engineering. We provide a review of research that examines the use of different media channels in software engineering from 1968 to the present day. We also provide preliminary results from a large survey with developers that actively use social media to understand how they communicate and collaborate, and to gain insights into the challenges they face. We find that while this particular population values social media, traditional channels, such as face-to-face communication, are still considered crucial. We synthesize findings from our historical review and survey to propose a roadmap for future research on this topic. Finally, we discuss implications for research methods as we argue that social media is poised to bring about a paradigm shift in software engineering research.
... In collaborative settings, awareness of others' emotions has been shown to enable users to respond accordingly and subsequently to achieve better results in collaborative work (García et al., 1999;Dullemond et al., 2013). This complements the knowledge in computer-supported cooperative work that awareness of significant information about others is beneficial in collaborative work settings (Gutwin & Greenberg, 2002). ...
Full-text available
Professional and lifelong learning are a necessity for workers. This is true both for re-skilling from disappearing jobs, as well as for staying current within a professional domain. AI-enabled scaffolding and just-in-time and situated learning in the workplace offer a new frontier for future impact of AIED. The hallmark of this community’s work has been i) data-driven design of learning technology and ii) machine-learning enabled personalized interventions. In both cases, data are the foundation of AIED research and data-related ethics are thus central to AIED research. In this paper we formulate a vision how AIED research could address data-related ethics issues in informal and situated professional learning. The foundation of our vision is a secondary analysis of five research cases that offer insights related to data-driven adaptive technologies for informal professional learning. We describe the encountered data-related ethics issues. In our interpretation, we have developed three themes: Firstly, in informal and situated professional learning, relevant data about professional learning – to be used as a basis for learning analytics and reflection or as a basis for adaptive systems - is not only about learners. Instead, due to the situatedness of learning, relevant data is also about others (colleagues, customers, clients) and other objects from the learner’s context. Such data may be private, proprietary, or both. Secondly, manual tracking comes with high learner control over data. Thirdly, learning is not necessarily a shared goal in informal professional learning settings. From an ethics perspective, this is particularly problematic as much data that would be relevant for use within learning technologies hasn’t been collected for the purposes of learning. These three themes translate into challenges for AIED research that need to be addressed in order to successfully investigate and develop AIED technology for informal and situated professional learning. As an outlook of this paper, we connect these challenges to ongoing research directions within AIED – natural language processing, socio-technical design, and scenario-based data collection - that might be leveraged and aimed towards addressing data-related ethics challenges.
... Recently, research on detection of emotions, politeness and sentiment in software engineering interactions has taken off [13], [16], [18], [20], [23], [24]. Except for Dullemond et al. [11], most of this work focuses on measuring the presence of some kind of conflict or negative feelings, with the aim of informing managers about these. Codes of conduct are a concrete tool to act on such information. ...
Conference Paper
Full-text available
Open source projects rely on collaboration of members from all around the world using web technologies like GitHub and Gerrit. This mixture of people with a wide range of backgrounds including minorities like women, ethnic minorities, and people with disabilities may increase the risk of offensive and destroying behaviours in the community, potentially leading affected project members to leave towards a more welcoming and friendly environment. To counter these effects, open source projects increasingly are turning to codes of conduct, in an attempt to promote their expectations and standards of ethical behaviour. In this first of its kind empirical study of codes of conduct in open source software projects, we investigated the role, scope and influence of codes of conduct through a mixture of quantitative and qualitative analysis, supported by interviews with practitioners. We found that the top codes of conduct are adopted by hundreds to thousands of projects, while all of them share 5 common dimensions.
... They used Latent Dirichlet Allocation to find the topics discussed in email and web discussions of students in the context of a class project, then use lexical sentimental analysis to obtain an average emotion score for each of the topics. Dullemond et al. [35] extended a microblogging tool with a happiness indicator, then deployed the tool across distributed teams in a company. Employees used the tool to share their message and emotions with their colleagues in order to stay more connected and be aware of each other while collaborating. ...
Full-text available
Software Engineering field has many goals, among them we can certainly deal with monitoring and controlling the development process in order to meet the business requirements of the released software artifact. Software engineers need to have empirical evidence that the development process and the overall quality of software artifacts is converging to the required features. Improving the development process's Effectiveness leads to higher productivity, meaning shorter time to market, but understanding or even measuring the software de- velopment process is an hard challenge. Modern software is the result of a complex process involving many stakeholders such as product owners, quality assurance teams, project manager and, above all, developers. All these stake- holders use complex software systems for managing development process, issue tracking, code versioning, release scheduling and many other aspect concerning software development. Tools for project management and issues/bugs tracking are becoming useful for governing the development process of Open Source soft- ware. Such tools simplify the communications process among developers and ensure the scalability of a project. The more information developers are able to exchange, the clearer are the goals, and the higher is the number of developers keen on joining and actively collaborating on a project. By analyzing data stored in such systems, researchers are able to study and address questions such as: Which are the factors able to impact the software productivity? Is it possible to improve software productivity shortening the time to market?. The present work addresses two major aspect of software development pro- cess: Effectiveness and Affectiveness. By analyzing data stored in project man- agement and in issue tracking system of Open Source Communities, we mea- sured the Effectiveness as the time required to resolve an issue and analyzed factors able to impact it.
... They used Latent Dirichlet Allocation to find the topics discussed in email and web discussions of students in the context of a class project, then use lexical sentimental analysis to obtain an average emotion score for each of the topics. Dullemond et al. [12] extended a microblogging tool with a happiness indicator, then deployed the tool across distributed teams in a company. Employees used the tool to share their message and emotions with their colleagues in order to stay more connected and be aware of each other while collaborating. ...
Full-text available
Software development is a collaborative activity in which developers interact to create and maintain a complex software system. Human collaboration inevitably evokes emotions like joy or sadness, which can affect the collaboration either positively or negatively, yet not much is known about the individual emotions and their role for software development stakeholders. In this study, we analyze whether development artifacts like issue reports carry any emotional information about software development. This is a first step towards verifying the feasibility of an automatic tool for emotion mining in software development artifacts: if humans cannot determine any emotion from a software artifact, neither can a tool. Analysis of the Apache Software Foundation issue tracking system shows that developers do express emotions (in particular gratitude, joy and sadness). However, the more context is provided about an issue report, the more human raters start to doubt and nuance their interpretation of emotions. More investigation is needed before building a fully automatic emotion mining tool.
Effective knowledge exchange among software developers is crucial for the competitive performance of their organizations. Today, the constant pressure on businesses to continually innovate and the increasing capability of information technologies to facilitate broader and more distributed communication are driving organizations to leverage social media tools to improve performance. These tools, which have changed the way we share knowledge, enable people to connect, communicate, and collaborate. Research on knowledge sharing via social media is still in its early phases, with a comprehensive overview of the literature yet to be completed. Thus, using a systematic literature review approach, this study aims to map the current literature on the topic in relation to software development. Furthermore, this study highlights the findings of former research and identifies gaps in the literature. The study offers several insights for researchers and practitioners and proposes a future research agenda to strengthen knowledge in the field.
Conference Paper
Recent innovations in social media have led to a paradigm shift in software development, with highly tuned participatory development cultures contributing to crowdsourced content and being supported by media that have become increasingly more social and transparent. Never before in the history of software development have we seen such rapid adoption of new tools by software developers. But there are many unanswered questions about the impact this tool adoption has on the quality of the software, the productivity and skills of the developers, the growth of projects and technologies developers contribute to, and how users can give feedback on and guide the software they use. Answering these questions is not trivial as this participatory development culture has become a virtual network of tightly coupled ecosystems consisting of developers, shared content and media channels. In our studies, we have found that combining research methods from the social sciences with data mining and software analytics to be the most promising in terms of revealing benefits and challenges from this adoption of social tools. In this talk, I share some of the findings from our studies, discuss the particular research methods we have used, and share our experiences from using those research methods. I also discuss how we as researchers leverage social tools and interact with the participatory development culture to assist with and help us gain feedback on our research. I close with a discussion about how other software engineering researchers could benefit from using social tools and the challenges they may face while doing so.
Throughout the history of software engineering, the human aspects have repeatedly been recognized as important. Even though research that investigates them has been growing in the past decade, these aspects should be more generally considered. The main objective of this study is to clarify the research area concerned with human aspects of software engineering and to create a common platform for future research. In order to meet the objective, we propose a definition of the research area behavioral software engineering (BSE) and present results from a systematic literature review based on the definition. The result indicates that there are knowledge gaps in the research area of behavioral software engineering and that earlier research has been focused on a few concepts, which have been applied to a limited number of software engineering areas. The individual studies have typically had a narrow perspective focusing on few concepts from a single unit of analysis. Further, the research has rarely been conducted in collaboration by researchers from both software engineering and social science. Altogether, this review can help put a broader set of human aspects higher on the agenda for future software engineering research and practice.
Conference Paper
Full-text available
Awareness of own and of other people's mood are prerequisite to a reflective learning process which allows people to consciously change their perception of, attitude towards and behaviour in future work situations. In this paper we investigate the usage and usefulness of the MoodMap App – an application for tracking own mood and creating awareness about the mood of team members in virtual meetings. Our study shows that especially the possibility to compare own mood to the mood of others' is perceived as useful and therefore enhances interpersonal communication in virtual team settings. Whilst users express an interest in tracking own mood, they need to relate their mood to the current context, and wish to receive feedback or other helpful input from the app in order to achieve propitious reflective learning.
Full-text available
Giant strides in information technology at the turn of the century may have unleashed unreachable goals. With the invention of groupware, people expect to communicate easily with each other and accomplish difficult work even though they are remotely located or rarely overlap in time. Major corporations launch global teams, expecting that technology will make "virtual collocation" possible. Federal research money encourages global science through the establishment of "collaboratories." We review over 10 years of field and laboratory investigations of collocated and noncollocated synchronous group collaborations. In particular, we compare collocated work with remote work as it is possible today and comment on the promise of remote work tomorrow. We focus on the sociotechnical conditions required for effective distance work and bring together the results with four key concepts: common ground, coupling of work, collaboration readiness, and collaboration technology readiness. Groups with high common ground and loosely coupled work, with readiness both for collaboration and collaboration technology, have a chance at succeeding with remote work. Deviations from each of these create strain on the relationships among teammates and require changes in the work or processes of collaboration to succeed. Often they do not succeed because distance still matters.
Conference Paper
Full-text available
Software engineering is by nature a highly collaborative activity. However, collaborating effectively in Global Software Engineering, in which team members are geographically, temporally and/or socio-culturally separated from each other, is more difficult. In a traditional co-located setting, one of the most important communication patterns is a (face-to-face) conversation. Technological solutions to have conversations in a distributed setting are commonly used, however overhearing conversations of others is not explicitly supported. In this paper we report on the evaluation of supporting overhearing conversations with technology in a distributed industrial setting. To do this we deployed a tool we developed with which it is possible to overhear Instant Messaging conversations in an international software development company. Based on this evaluation we report lessons learned and conclude with the most important findings of this study.
Conference Paper
Full-text available
Affective aspects during collaboration can be exploited as triggers for reflection, yet current tools usually ignore these aspects. In this paper, we present a set of design choices to inform the design of systems towards enabling mood awareness in collaborative work settings like meetings or conferences. Design choices have served to outline and implement a collection of prototypes, including two input interfaces and three visualizations, which have been evaluated during an important project meeting over three days. Our results show that (a) the user acceptance of capturing mood is high (b) the aggregated mood values are related to the work process and (c) aggregated moods can influence the individual by creating awareness of others. Further discussion on the impact of the different design choices shows promising venues to improve mood awareness support.
Conference Paper
Today's generation of software developers frequently make use of social media, either as an adjunct or integrated into a wide range of tools ranging from code editors and issue trackers, to IDEs and web-based portals. The role of social media usage in software engineering is not well understood, and yet the use of these mechanisms influences software development practices. In this position paper, we advocate for research that strives to understand the benefits, risks and limitations of using social media in software development at the team, project and community levels. Guided by the implications of current tools and social media features, we propose a set of pertinent research questions around community involvement, project coordination and management, as well as individual software development activities. Answers to these questions will guide future software engineering tool innovations and software development team practices.
Conference Paper
Cooperative processes are strongly influenced by their context. Individual and group experience, work practices, and the organisational setting are variable factors that help shape a cooperative work context. In an electronic environment the variability and dynamic nature of cooperative processes has to be taken into account. This suggests a requirement for configurability of basic cooperative functionality. In this paper, a concept for the development of a generic CSCW platform is proposed, which offers the possibility to configure its basic cooperation support functionality. Mediating objects are the key feature which enables the tailorability of functionality to the cooperative setting at run-time.
Conference Paper
This is a case study about the early adoption and use of micro-blogging in a Fortune 500 company. The study used several independent data sources: five months of empirical micro-blogging data, user demographic information from corporate HR records, a web based survey, and targeted interviews. The results revealed that users vary in their posting activities, reading behaviors, and perceived benefits. The analysis also identified barriers to adoption, such as the noise-to-value ratio paradoxes. The findings can help both practitioners and scholars build an initial understanding of how knowledge workers are likely to use micro-blogging in the enterprise.